Author: "GIPSA Pôle Sciences des Données (GIPSA-PSD)" - Searchworks@Jio Institute Digital Library Search Results

Your search keyword '"GIPSA Pôle Sciences des Données (GIPSA-PSD)"' showing total 370 results

Start Over Author "GIPSA Pôle Sciences des Données (GIPSA-PSD)"

370 results on '"GIPSA Pôle Sciences des Données (GIPSA-PSD)"'

1. Classification d'images hyperspectrales basée sur la morphologie mathématique et la décomposition tensorielle

Author: Jouni, Mohamad, Dalla Mura, Mauro, Comon, Pierre, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), GIPSA Pôle Géométrie, Apprentissage, Information et Algorithmes (GIPSA-GAIA), Tokyo Tech World Research Hub Initiative (WRHI), School of Computing, Tokyo Institute of Technology, Tokyo, SIGMAPHY (GIPSA-SIGMAPHY), Département Images et Signal (GIPSA-DIS), Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-GIPSA Pôle Sciences des Données (GIPSA-PSD), Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA), CICS (GIPSA-CICS), Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-GIPSA Pôle Sciences des Données (GIPSA-PSD), and Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Université Pierre Mendès France - Grenoble 2 (UPMF)-Université Stendhal - Grenoble 3-Université Joseph Fourier - Grenoble 1 (UJF)-Institut Polytechnique de Grenoble - Grenoble Institute of Technology-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)
Subjects: Ten- sor Decomposition, Computer Science::Computer Vision and Pattern Recognition, Attribute Profiles, [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], Tensor Decomposition, Physics::Accelerator Physics, Scene Classification MSC: 15A69, Remote Sensing Image, Mathematical Morphology, Hyperspectral Imagery, Scene Classification, 15A69, 94A08, 94A08
Abstract: International audience; Hyperspectral Image (HSI) classification refers to classifying hyperspectral data into features, where labels are given to pixels sharing the same features, distinguishing the present materials of the scene from one another. Naturally a HSI acquires spectral features of pixels, but spatial features based on neighborhood information are also important, which results in the problem of spectral-spatial classification. There are various ways to account to spatial information, one of which is through Mathematical Morphology, which is explored in this work. A HSI is a third-order data block, and building new spatial diversities may increase this order. In many cases, since pixel-wise classification requires a matrix of pixels and features, HSI data are reshaped as matrices which causes high dimensionality and ignores the multi-modal structure of the features. This work deals with HSI classification by modeling the data as tensors of high order. More precisely, multi-modal hyperspectral data is built and dealt with using tensor Canonical Polyadic (CP) decomposition. Experiments on real HSI show the effectiveness of the CP decomposition as a candidate for classification thanks to its properties of representing the pixel data in a matrix compact form with a low dimensional feature space while maintaining the multi-modality of the data.
Published: 2020

2. SHCNet: A semi-supervised hypergraph convolutional networks based on relevant feature selection for hyperspectral image classification

Author: Akrem Sellami, Mohamed Farah, Mauro Dalla Mura, Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 (CRIStAL), Centrale Lille-Université de Lille-Centre National de la Recherche Scientifique (CNRS), Laboratoire de recherche en Génie Logiciel, Applications distribuées, Systèmes décisionnels et Imagerie intelligente [Manouba] (RIADI), École Nationale des Sciences de l'Informatique [Manouba] (ENSI), Université de la Manouba [Tunisie] (UMA)-Université de la Manouba [Tunisie] (UMA), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Institut Universitaire de France (IUF), and Ministère de l'Education nationale, de l’Enseignement supérieur et de la Recherche (M.E.N.E.S.R.)
Subjects: [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Artificial Intelligence, hyperspectral image classification, Signal Processing, Unsupervised feature selection, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], Computer Vision and Pattern Recognition, hypergraph convolutional network, Software, dimensionality reduction, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]
Abstract: International audience; Hyperspectral imagery classification is a challenging task due to the large number of spectral bands, and low number of labeled samples. To overcome these issues, we propose a novel approach for hyperspectral image classification based on feature selection and semi-supervised hypergraph convolutional network working with small number of labeled samples. Firstly, we propose a new unsupervised feature selection method based on an information theoretic criterion. Relevant spectral features are automatically selected while preserving the physical properties of hyperspectral data. Secondly, we construct a spectro-spatial hypergraph in order to represent the complex relationships between pixels. Finally, we propose a semi-supervised hypergraph convolutional network which integrates local vertex features and hypergraph topology in the convolutional layers. The aim of this step is to preserve the spectro-spatial features and to cope with the high correlation between hypernodes during classification. The main advantage of the proposed approach is to allow the automatic selection of relevant spectral bands while preserving the spatial and spectral features. In addition, by accounting for the relationship between pixels leads to improved classification results even when the number of labeled samples is low. Experiments are conducted on two real hyperspectral images show that the proposed approach reaches competitive good performances, and achieves better classification performances compared to state-of-the-art methods.
Published: 2023

3. Hyperspectral Image Super-Resolution via Deep Spatiospectral Attention Convolutional Neural Networks

Author: Gemine Vivone, Jocelyn Chanussot, Jin-Fan Hu, Ting-Zhu Huang, Tai-Xiang Jiang, Liang-Jian Deng, University of Electronic Science and Technology of China [Chengdu] (UESTC), Southwestern University of Finance and Economics [Chengdu, China], Institute of Methodologies for Environmental Analysis of the National Research Council (IMAA), National Research Council of Italy | Consiglio Nazionale delle Ricerche (CNR), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Source code, Hyperspectral imaging, Mean squared error, Computer Networks and Communications, Computer science, media_common.quotation_subject, Multispectral image, Tensors, Superresolution, Convolutional neural network, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Artificial Intelligence, Robustness (computer science), Computer architecture, Image resolution, media_common, Spatial resolution, Network architecture, Training data, Learning systems, business.industry, Pattern recognition, Computer Science Applications, Artificial intelligence, business, Software
Abstract: International audience; Hyperspectral images (HSIs) are of crucial importance in order to better understand features from a large number of spectral channels. Restricted by its inner imaging mechanism, the spatial resolution is often limited for HSIs. To alleviate this issue, in this work, we propose a simple and efficient architecture of deep convolutional neural networks to fuse a low-resolution HSI (LR-HSI) and a high-resolution multispectral image (HR-MSI), yielding a high-resolution HSI (HR-HSI). The network is designed to preserve both spatial and spectral information thanks to a new architecture based on: 1) the use of the LR-HSI at the HR-MSI’s scale to get an output with satisfied spectral preservation and 2) the application of the attention and pixelShuffle modules to extract information, aiming to output high-quality spatial details. Finally, a plain mean squared error loss function is used to measure the performance during the training. Extensive experiments demonstrate that the proposed network architecture achieves the best performance (both qualitatively and quantitatively) compared with recent state-of-the-art HSI super-resolution approaches. Moreover, other significant advantages can be pointed out by the use of the proposed approach, such as a better network generalization ability, a limited computational burden, and the robustness with respect to the number of training samples. Please find the source code and pretrained models from https://liangjiandeng.github.io/Projects_Res/HSRnet_2021tnnls.html .
Published: 2022

4. Event-based neural learning for quadrotor control

Author: Carvalho, Estéban, Susbielle, Pierre, Marchand, Nicolas, Hably, Ahmad, Dibangoye, Jilles, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA), CITI Centre of Innovation in Telecommunications and Integration of services (CITI), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National de Recherche en Informatique et en Automatique (Inria), GIPSA-Services (GIPSA-Services), Université de Lyon, Robots coopératifs et adaptés à la présence humaine en environnements (CHROMA), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-CITI Centre of Innovation in Telecommunications and Integration of services (CITI), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-Inria Lyon, Institut National de Recherche en Informatique et en Automatique (Inria), ANR-11-LABX-0025,PERSYVAL-lab,Systemes et Algorithmes Pervasifs au confluent des mondes physique et numérique(2011), ANR-10-EQPX-0044,ROBOTEX,Réseau national de plateformes robotiques d'excellence(2010), and ANR-21-ESRE-0015,TIRREX,Infrastructure technologique pour la recherche d'excellence en robotique(2021)
Subjects: event-based control, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Quadrotor, [INFO.INFO-AU]Computer Science [cs]/Automatic Control Engineering, online learning, deep neural network, trajectory tracking, [SPI.AUTO]Engineering Sciences [physics]/Automatic
Abstract: International audience; The design of a simple and adaptive flight controller is a real challenge in aerial robotics. A simple flight controller often generates a poor flight tracking performance. Furthermore, adaptive algorithms might be costly in time and resources or deep learning based methods may cause instability problems, for instance in presence of disturbances. In this paper, we propose an event-based neural learning control strategy that combines the use of a standard cascaded flight controller enhanced by a deep neural network that learns the disturbances in order to improve the tracking performance. The strategy relies on two events: one allowing the improvement of tracking errors and the second to ensure closed-loop system stability. After a validation of the proposed strategy in a ROS/Gazebo simulation environment, its effectiveness is confirmed in real experiments in the presence of wind disturbance.
Published: 2023

5. Slow slip detection with deep learning in multi-station raw geodetic time series validated against tremors in Cascadia

Author: Costantino, Giuseppe, Giffard-Roisin, Sophie, Radiguet, Mathilde, Dalla Mura, Mauro, Marsan, David, Socquet, Anne, Institut des Sciences de la Terre (ISTerre), Institut national des sciences de l'Univers (INSU - CNRS)-Institut de recherche pour le développement [IRD] : UR219-Université Savoie Mont Blanc (USMB [Université de Savoie] [Université de Chambéry])-Centre National de la Recherche Scientifique (CNRS)-Université Gustave Eiffel-Université Grenoble Alpes (UGA), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and Université Grenoble Alpes (UGA)
Subjects: GNSS, neural network, GPS, deformation, deep learning, Cascadia, attention-based, [PHYS.PHYS.PHYS-GEO-PH]Physics [physics]/Physics [physics]/Geophysics [physics.geo-ph], tremor, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], slow slip events, transient, classification, geodesy, transformer, multi-station, earthquakes, subduction
Abstract: International audience; Slow slip events (SSEs) originate from a slow slippage on faults that lasts from a few days to years. A systematic and complete mapping of SSEs is key to characterizing the slip spectrum and understanding its link with coeval seismological signals. Yet, SSE catalogues are sparse and usually remain limited to the largest events, because the deformation transients are often concealed in the noise of the geodetic data. Here we present the first multi-station deep learning SSE detector applied blindly to multiple raw geodetic time series. Its power lies in an ultra-realistic synthetic training set, and in the combination of convolutional and attentionbased neural networks. Applied to real data in Cascadia over the period 2007-2022, it detects 78 SSEs, that compare well to existing independent benchmarks: 87.5% of previously catalogued SSEs are retrieved, each detection falling within a peak of tremor activity. Our method also provides useful proxies on the SSE duration and may help illuminate relationships between tremor chatter and the nucleation of the slow rupture. We find an average day-long time lag between the slow deformation and the tremor chatter both at a global-and local-temporal scale, suggesting that slow slip may drive the rupture of nearby small asperities.
Published: 2023

6. Privacy protection control for mobile apps users

Author: Sophie Cerf, Bogdan Robu, Nicolas Marchand, Sara Bouchenak, Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL], GIPSA - Modelling and Optimal Decision for Uncertain Systems [GIPSA-MODUS], GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing [GIPSA-COPERNIC], Institut National des Sciences Appliquées de Lyon [INSA Lyon], Laboratoire d'InfoRmatique en Image et Systèmes d'information [LIRIS], Distribution, Recherche d'Information et Mobilité [DRIM], Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 (CRIStAL), Centrale Lille-Université de Lille-Centre National de la Recherche Scientifique (CNRS), GIPSA - Modelling and Optimal Decision for Uncertain Systems (GIPSA-MODUS), GIPSA Pôle Automatique et Diagnostic (GIPSA-PAD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA), Laboratoire d'InfoRmatique en Image et Systèmes d'information (LIRIS), Université Lumière - Lyon 2 (UL2)-École Centrale de Lyon (ECL), Université de Lyon-Université de Lyon-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Centre National de la Recherche Scientifique (CNRS), Distribution, Recherche d'Information et Mobilité (DRIM), and Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Centre National de la Recherche Scientifique (CNRS)-Université Lumière - Lyon 2 (UL2)-École Centrale de Lyon (ECL)
Subjects: Control and Systems Engineering, Applied Mathematics, control of computing systems location privacy differential-privacy modeling Sampled-Data Control, [INFO]Computer Science [cs], Electrical and Electronic Engineering, Computer Science Applications, [SPI.AUTO]Engineering Sciences [physics]/Automatic
Abstract: International audience; Predominant in today society, mobile apps are rising as promising application systems for automatic control. An app can be viewed as a plant, processing input signals (queries, phone data, etc.) and generating outputs (such as a service or an answer). Guaranteeing that the app complies with a desired behavior is a major safety challenge. This work focuses on privacy issues for geolocated mobile apps. Many applications use the location data to provide a service (e.g., navigation, fitness) or to improve it (e.g., weather forecast, social media). This gain in service utility comes at the cost of personal data sharing. Such threat to user privacy can be leveraged by protection mechanisms, e.g., addition of noise to the location data. However, state-of-the-art techniques still lack means of ensuring both data utility and privacy in a dynamics utilization context. This paper presents the first non-linear analytical modeling followed by a control formulation for regulating the privacy level in a mobile app. The privacy is sensed using the well established notion of Point of Interest. Through modeling, we highlight the control challenges, namely the non-linearity and time-variance of the plant, its high sensibility to noise and the impact of the user's mobility pattern-seen a disturbance. A controller is designed, combining feedback with anticipation action. Evaluation is performed using mobility records from two real-world multi-users datasets. Our approach enables, with a unique and universal tuning, to robustly meet privacy objectives with preserved utility and negligible computational overhead. Control algorithm, experimental evaluation and analysis scripts are available online for reproducibility.
Published: 2023

7. Leveraging Neural Koopman Operators to Learn Continuous Representations of Dynamical Systems from Scarce Data

Author: Frion, Anthony, Drumetz, Lucas, Dalla Mura, Mauro, Tochon, Guillaume, Aissa El Bey, Abdeldjalil, Equipe Observations Signal & Environnement (Lab-STICC_OSE), Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance (Lab-STICC), École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS)-Université Bretagne Loire (UBL)-IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-École Nationale d'Ingénieurs de Brest (ENIB)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Institut Mines-Télécom [Paris] (IMT)-Centre National de la Recherche Scientifique (CNRS)-Université Bretagne Loire (UBL)-IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT), Département Mathematical and Electrical Engineering (IMT Atlantique - MEE), IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT), Océan Dynamique Observations Analyse (ODYSSEY), Université de Bretagne Occidentale - UFR Sciences et Techniques (UBO UFR ST), Université de Brest (UBO)-Université de Brest (UBO)-Université de Rennes (UR)-Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER)-Inria Rennes – Bretagne Atlantique, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-IMT Atlantique (IMT Atlantique), Institut Universitaire de France (IUF), Ministère de l'Education nationale, de l’Enseignement supérieur et de la Recherche (M.E.N.E.S.R.), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Laboratoire de Recherche et de Développement de l'EPITA (LRDE), Ecole Pour l'Informatique et les Techniques Avancées (EPITA), Equipe Communication System Design (Lab-STICC_COSYDE), ANR-21-CE48-0005,LEMONADE,Apprentissage et modélisation de la dynamique spectrale de séries temporelles d'images satellitaires(2021), Université de Brest (UBO)-Université de Brest (UBO)-Université de Rennes 1 (UR1), and Université de Rennes (UNIV-RENNES)-Université de Rennes (UNIV-RENNES)-Institut Français de Recherche pour l'Exploitation de la Mer (IFREMER)-Inria Rennes – Bretagne Atlantique
Subjects: FOS: Computer and information sciences, Computer Science - Machine Learning, [MATH.MATH-DS]Mathematics [math]/Dynamical Systems [math.DS], Continuous dynamical systems, Dynamical Systems (math.DS), [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI], Machine Learning (cs.LG), Data assimilation, FOS: Mathematics, [INFO]Computer Science [cs], [MATH]Mathematics [math], Mathematics - Dynamical Systems, Learning dynamical systems, Koopman operator, Forecasting
Abstract: International audience; Over the last few years, several works have proposed deep learning architectures to learn dynamical systems from observation data with no or little knowledge of the underlying physics. A line of work relies on learning representations where the dynamics of the underlying phenomenon can be described by a linear operator, based on the Koopman operator theory. However, despite being able to provide reliable long-term predictions for some dynamical systems in ideal situations, the methods proposed so far have limitations, such as requiring to discretize intrinsically continuous dynamical systems, leading to data loss, especially when handling incomplete or sparsely sampled data. Here, we propose a new deep Koopman framework that represents dynamics in an intrinsically continuous way, leading to better performance on limited training data, as exemplified on several datasets arising from dynamical systems.
Published: 2023

8. Simulation of Spinal Muscle Control in Human Gait Using OpenSim

Author: Julien Frère, Patrick HENAFF, Andrii Shachykov, Analysis and modeling of neural systems by a system neuroscience approach (NEUROSYS), Inria Nancy - Grand Est, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Department of Complex Systems, Artificial Intelligence & Robotics (LORIA - AIS), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), GIPSA - Analyse et Modification de l'homme en mouvement : biomécanique, cognition, vocologie (GIPSA-MOVE), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and Université Grenoble Alpes (UGA)
Subjects: CPG, [SDV]Life Sciences [q-bio], human walking, OpenSim, musculoskeletal model
Abstract: International audience; This paper presents a neuro-musculoskeletal simulation approach to human gait based on an original model of central pattern generator and the muscle synergy approach. The controller, shown here, aims at simplifying movement control by reducing the number of parameters to optimize. The model of the simplified motor coordination chain was built, from the midbrain through spinal neurons to the muscles allowing of simulating some neural and muscular values of gait, hard to obtain otherwise. The work also includes a bio-inspired reflexbased balance controller designed from knowledge of spinal gait. The used anatomic musculoskeletal model has been implemented in OpenSim platform. Time simulations show the ability of the platform to be able to produce different gait patterns including nominal and disturbed. Both gaits are controlled by the same bio-inspired closed-loop CPG-and reflex-based circuitries. After 2 second standing phase, the model was able to walk for about 5 meters with 10 steps. 2 steps were disrupted in the second simulation.
Published: 2022

9. An Optimization Procedure for Robust Regression-Based Pansharpening

Author: Marco Carpentiero, Gemine Vivone, Rocco Restaino, Paolo Addesso, Jocelyn Chanussot, University of Salerno (UNISA), Institute of Methodologies for Environmental Analysis of the National Research Council (IMAA), National Research Council of Italy | Consiglio Nazionale delle Ricerche (CNR), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Optimization, regression-based data fusion, Sensors, pansharpening, Maximum likelihood estimation, Pansharpening, Estimation, Laplace equations, Electronic mail, Image fusion, pyramidal decomposition, remote sensing, General Earth and Planetary Sciences, Electrical and Electronic Engineering, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Model-based approaches to pansharpening still constitute a class of widely employed methods, thanks to their straightforward applicability to many problems, dispensing the user from time-consuming training phases. The injection scheme based on an accurate estimation (exploiting regression) of the relationship between the details contained in the panchromatic (PAN) image and those required for the enhancement of the multispectral (MS) image represents the most updated approach to this problem, being characterized by both theoretical and practical optimality. We elaborated on this scheme by designing a procedure for estimating the key parameters required for the optimal setting of such a regression-based approach. We tested this approach on several datasets acquired by the WorldView satellites comparing the proposed approach with a benchmark consisting of some state-of-the-art pansharpening methods.
Published: 2022

10. N-Cluster Loss and Hard Sample Generative Deep Metric Learning for PolSAR Image Classification

Author: Licheng Jiao, Bo Ren, Biao Hou, Shuang Wang, Chen Yang, Jocelyn Chanussot, Yue Hu, Xidian University, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Measurement, Discriminator, Similarity (geometry), Contextual image classification, business.industry, Computer science, Feature vector, Deep learning, Pattern recognition, Sample (statistics), Scattering, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Task analysis, Metric (mathematics), Feature extraction, Training, General Earth and Planetary Sciences, Artificial intelligence, Electrical and Electronic Engineering, Convergence, business, Focus (optics), Data mining
Abstract: International audience; Deep learning works normally in PolSAR image classification because the complex terrain scattering characteristic results in large intraclass differences and high interclass similarity. Deep metric learning (DML) aims to make the features keep a closer intraclass and a farther interclass distance. Therefore, we introduce DML and then propose an N-cluster generative adversarial net (N-cluster GAN) framework for PolSAR image classification. However, existing DML losses mainly focus on the relationship between individual samples in feature space. Hence, we propose N-cluster loss that pays more attention to the overall structure of all samples. Meanwhile, traditional hard negative sample mining methods occupy lots of computational resources. In addition, the hard level of the negative samples will affect the model’s performance. Therefore, we explore a new method based on a GAN framework to replace the sample mining. Positive N-cluster loss is added to the discriminator ( D ), and a negative one is added to the generator ( G ). In this way, D will possess better classification ability, and G can produce hard negative samples for D . Then, the hard level of the generated negative samples will change with the discrimination of D , which is appropriate for the proposed model. N-cluster loss can be directly calculated through the extracted features rather than redundant data preparation. The proposed model is verified on four PolSAR datasets from two aspects of the loss function and negative samples mining. Then, it achieves competitive performance compared with state-of-the-art algorithms.
Published: 2022

11. PDFL: Polarimetric Decomposition Feature Learning via Deep Autoencoder

Author: Chen Yang, Biao Hou, Qian Wu, Bo Ren, Jocelyn Chanussot, Shuang Wang, Licheng Jiao, Xidian University, Air Force Engineering University [Xi'an] (AFEU), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Scattering, Mathematical models, Matrix decomposition, Covariance matrices, Decoding, Data models, General Earth and Planetary Sciences, Electrical and Electronic Engineering, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Representation learning
Abstract: International audience; Model-based polarimetric target decomposition (TD) generally solves scattering components and parameters under a preset decomposition base; then, decomposition features are also obtained. However, a preset base could not be adjusted according to different scenes. Furthermore, solving the polarimetric parameters needs exploring additional information or considering limiting conditions to build equations, which is hard and easily brings negative effects into decomposition features. To this end, we regard the TD as a process of learning decomposition base and features by deep learning. Then, the polarimetric decomposition feature learning (PDFL) model is proposed in this article. Strictly, this model is not an incoherent TD method but a learning-based method. It does not need to construct the parameter solution equations or fixed base. Then, the decomposition base and feature can be adaptively learned according to the scattering characteristics of the current dataset. Due to the characteristics of unsupervised reconstruction, the deep autoencoder (DAE) is used as the model foundation. Then, some adjustments and constraints are utilized to make the DAE fit closely with TD. The encoder extracts the latent vector from polarimetric synthetic aperture radar (PolSAR) data, and then, the decoder reconstructs pseudodata on this latent vector. The reconstruction can be regarded as the inverse process of TD, so the base matrix of the decoder and the latent vector indicate the learned decomposition base and features when the model converges. The effectiveness of PDFL is verified on simulated and real PolSAR datasets. Compared with representative algorithms, the proposed model gains more discriminative features and achieves competitive performance on terrain classification and segmentation tasks.
Published: 2022

12. Reconstruction Error-Based Decomposition Feature Selection for PolSAR Image

Author: Biao Hou, Chen Yang, Xianpeng Guo, Licheng Jiao, Bo Ren, Jocelyn Chanussot, Shuang Wang, Xidian University, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: business.industry, Computer science, Data models, Synthetic aperture radar, Feature selection, Pattern recognition, Image (mathematics), Scattering, Reconstruction error, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Task analysis, Image reconstruction, Decomposition (computer science), Feature extraction, Training, General Earth and Planetary Sciences, Artificial intelligence, Electrical and Electronic Engineering, business
Abstract: International audience; Target decomposition features are the cornerstone of subsequent analyses for PolSAR images. Generally, adopting single or several decomposition algorithms limits the representation ability for original terrain characteristics. Using all the existing decomposition features, however, will definitely increase computational complexity. Besides, some features even have a negative effect on the following tasks. To address these problems, a sparse variational autoencoder feature selection framework (SVAE-FS) is proposed in this article. In detail, the encoder transforms the original feature set into latent space and then decoder reconstructs the corresponding pseudo set on this latent space. Similarly, a pseudo subset is subsequently obtained by the SVAE. The discrepancy, namely reconstruction error, between the pseudo set and the pseudo subset is taken as an evaluation criterion which reflects the feature representation ability of pseudo subset. Sparse constraint in the encoder makes the representative features stand out. Meanwhile, the linear feature transformation layer of the encoder enables the SVAE to evaluate different scale subsets without repeated training. Finally, a greedy selection approach with search scale K is proposed to find the suboptimal subset. This procedure not only reduces time consumption, but also ensures the performance of the subset. The selected features are analyzed on four real PolSAR datasets according to the terrain scattering characteristics. Furthermore, these features have achieved competitive performance on three PolSAR image tasks.
Published: 2022

13. Convolutional Neural Networks for Multimodal Remote Sensing Data Classification

Author: Jocelyn Chanussot, Danfeng Hong, Xin Wu, Beijing Institute of Technology (BIT), Chinese Academy of Sciences [Beijing] (CAS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Synthetic aperture radar, Hyperspectral imaging, business.industry, Computer science, Deep learning, Data classification, Network architecture, Ranging, Convolutional neural network, Laser radar, Lidar, Modal, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Task analysis, Feature extraction, General Earth and Planetary Sciences, Convolutional neural networks, Artificial intelligence, Electrical and Electronic Engineering, business, Remote sensing
Abstract: International audience; In recent years, enormous research has been made to improve the classification performance of single-modal remote sensing (RS) data. However, with the ever-growing availability of RS data acquired from satellite or airborne platforms, simultaneous processing and analysis of multimodal RS data pose a new challenge to researchers in the RS community. To this end, we propose a deep-learning-based new framework for multimodal RS data classification, where convolutional neural networks (CNNs) are taken as a backbone with an advanced cross-channel reconstruction module, called CCR-Net. As the name suggests, CCR-Net learns more compact fusion representations of different RS data sources by the means of the reconstruction strategy across modalities that can mutually exchange information in a more effective way. Extensive experiments conducted on two multimodal RS datasets, including hyperspectral (HS) and light detection and ranging (LiDAR) data, i.e., the Houston2013 dataset, and HS and synthetic aperture radar (SAR) data, i.e., the Berlin dataset, demonstrate the effectiveness and superiority of the proposed CCR-Net in comparison with several state-of-the-art multimodal RS data classification methods. The codes will be openly and freely available at https://github.com/danfenghong/IEEE_TGRS_CCR-Net for the sake of reproducibility.
Published: 2022

14. Predictive and Symbolic Control: Performance and Safety for Non-linear Systems

Author: Azaki, Zakeye, Girard, Antoine, Olaru, Sorin, Girard, Antoine, Towards programmable cyber-physical systems: a symbolic control approach - PROCSYS - - H2020 Pilier ERC2017-09-01 - 2022-08-31 - 725144 - VALID, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Laboratoire des signaux et systèmes (L2S), CentraleSupélec-Université Paris-Saclay-Centre National de la Recherche Scientifique (CNRS), and European Project: 725144,H2020 Pilier ERC,PROCSYS(2017)
Subjects: [INFO.INFO-SY] Computer Science [cs]/Systems and Control [cs.SY], Recursive feasibility, [SPI.AUTO] Engineering Sciences [physics]/Automatic, Control and Systems Engineering, Non-linear systems, Controlled invariant set, [INFO.INFO-SY]Computer Science [cs]/Systems and Control [cs.SY], [MATH.MATH-OC] Mathematics [math]/Optimization and Control [math.OC], [MATH.MATH-OC]Mathematics [math]/Optimization and Control [math.OC], Model predictive control, Symbolic control, Terminal constraints, [SPI.AUTO]Engineering Sciences [physics]/Automatic
Abstract: International audience; Although Model Predictive Control (MPC) has proven its e ciency in the process operation and it is known for its high performance, it also su ers from design limitation. One of the fundamental issues for such an optimization-based technique is the di culty to guarantee recursive feasibility in the absence of terminal constraints, or alternatively, the complexity of design when a certain basin of a raction or a controlled invariant set needs to be certi ed for the closed loop in the most general nonlinear se ing. Based on symbolic control techniques, this paper proposes a simple and guaranteed solution for such problems. As a main result, a Symbolically guided Model Predictive Control scheme is developed. is controller is an improved version of the generic MPC approach such that the recursive feasibility is guaranteed through appending time-varying terminal constraints, carefully designed using the symbolic control approach, to the optimization problem of the original MPC formulation.
Published: 2022

15. CyCU-Net: Cycle-Consistency Unmixing Network by Learning Cascaded Autoencoders

Author: Jocelyn Chanussot, Zhu Han, Lianru Gao, Danfeng Hong, Bing Zhang, Chinese Academy of Sciences [Beijing] (CAS), German Aerospace Center (DLR), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: remote sensing (RS), business.industry, Computer science, Deep learning, 0211 other engineering and technologies, Process (computing), Hyperspectral imaging, Pattern recognition, 02 engineering and technology, cycle consistency, Net (mathematics), Autoencoder, Cascaded autoencoders (AEs), hyperspectral unmixing (HU), Term (time), self-perception, Consistency (database systems), deep learning (DL), [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Curve fitting, General Earth and Planetary Sciences, Artificial intelligence, Electrical and Electronic Engineering, business, 021101 geological & geomatics engineering
Abstract: International audience; In recent years, deep learning (DL) has attracted increasing attention in hyperspectral unmixing (HU) applications due to its powerful learning and data fitting ability. The autoencoder (AE) framework, as an unmixing baseline network, achieves good performance in HU by automatically learning low-dimensional embeddings and reconstructing data. Nevertheless, the conventional AE-based architecture, which focuses more on the pixel-level reconstruction loss, tends to lose some significant detailed information of certain materials (e.g., material-related properties) in the reconstruction process. Therefore, inspired by the perception mechanism, we propose a cycle-consistency unmixing network, called CyCU-Net, by learning two cascaded AEs in an end-to-end fashion, to enhance the unmixing performance more effectively. CyCU-Net is capable of reducing the detailed and material-related information loss in the process of reconstruction by relaxing the original pixel-level reconstruction assumption to cycle consistency dominated by the cascaded AEs. More specifically, cycle consistency can be achieved by a newly proposed self-perception loss, which consists of two spectral reconstruction terms and one abundance reconstruction term. By taking advantage of the self-perception loss in the network, the high-level semantic information can be well preserved in the unmixing process. Moreover, we investigate the performance gain of CyCU-Net with extensive ablation studies. Experimental results on one synthetic and three real hyperspectral data sets demonstrate the effectiveness and competitiveness of the proposed CyCU-Net in comparison with several state-of-the-art unmixing algorithms.
Published: 2022

16. Bayesian Unmixing of Hyperspectral Image Sequence With Composite Priors for Abundance and Endmember Variability

Author: Qian Du, Jocelyn Chanussot, Hongyi Liu, Zhihui Wei, Zebin Wu, Youkang Lu, Nanjing University of Science and Technology (NJUST), Mississippi State University [Mississippi], GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), and Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )
Subjects: Endmember, Hyperspectral imaging, Posterior probability, Bayesian probability, Principal component analysis, Statistics::Machine Learning, symbols.namesake, Image sequences, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Abundance (ecology), Prior probability, Electrical and Electronic Engineering, Lighting, Mathematics, business.industry, Markov processes, Monte Carlo methods, Pattern recognition, Markov chain Monte Carlo, Bayes methods, Computer Science::Computer Vision and Pattern Recognition, symbols, General Earth and Planetary Sciences, Artificial intelligence, Likelihood function, business
Abstract: International audience; A hyperspectral image sequence can be obtained at different time in the same region from a hyperspectral sensor. The environmental change usually leads to variation in endmember reflectance, which has an important influence on unmixing process. In this article, a Bayesian unmixing model considering spectral variability for hyperspectral sequence is proposed, in which composite prior distributions of abundance and endmember variability are developed. The abundance priors consider the continuity of abundance in the temporal and spatial domains, simultaneously. Specifically, in the spatial domain, a data-adaptive variance of the abundance prior distribution is put forward based on local spatial difference. Moreover, the priors of endmember variability in temporal continuity and spectral smoothness are also exploited. Finally, a joint posterior distribution is obtained by the likelihood function and the parameter prior distributions, which can be calculated by the Markov chain Monte Carlo (MCMC) algorithm. Experiments on synthetic and real data sets demonstrate the effectiveness of the proposed approach in terms of abundance, endmember, and its variability estimation accuracy.
Published: 2022

17. Hyperspectral Pansharpening With Adaptive Feature Modulation-Based Detail Injection Network

Author: Yunsong Li, Yuxuan Zheng, Jiaojiao Li, Rui Song, Jocelyn Chanussot, Xidian University, University College of London [London] (UCL), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Modulation, Spatial resolution, Image resolution, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Feature extraction, General Earth and Planetary Sciences, Pansharpening, Electrical and Electronic Engineering, Data mining, Convolution
Abstract: International audience; Recently, deep learning-based methodologies have attained unprecedented performance in hyperspectral (HS) pansharpening, which aims to improve the spatial quality of HS images (HSIs) by making use of details extracted from the high-resolution panchromatic (HR-PAN) image. However, it remains challenging to incorporate the details into the pansharpened image effectively, while alleviating the spectral distortion simultaneously. To tackle this problem, in this article, we propose an adaptive feature modulation-based detail injection network (AFM-DIN) for HS pansharpening, which mainly consists of four phases: high-frequency details generation of the HR-PAN image, multiscale feature extraction of the upsampled HSI, AFM-based detail injection, and reconstruction of the HR-HSI. First, a novel octave convolution unit is employed to decompose the HR-PAN image into high and low frequencies, and then merge the high-frequency features together to generate the comprehensive PAN-details. Second, the spatial and spectral separable 3-D convolution units with multiple kernel sizes are designed to extract multiscale features of the upsampled HSI in a computationally efficient manner. Subsequently, by taking the critical PAN-details as prior, the proposed AFM module is able to not only incorporate the detail information effectively, but also adjust the injected details adaptively to ensure the spectral fidelity. Finally, the anticipated HR-HSI is obtained through adding the upsampled HSI to the predicted HSI-details reconstructed from informative modulated features. Extensive comparison experiments with several state-of-the-arts conducted on simulated and real HS data sets demonstrate that our proposed AFM-DIN can achieve superior pansharpening accuracy in both spatial and spectral aspects.
Published: 2022

18. CNN-Based Hyperspectral Pansharpening With Arbitrary Resolution

Author: Lin He, Jiawei Zhu, Jun Li, Antonio Plaza, Jocelyn Chanussot, Zhuliang Yu, South China University of Technology [Guangzhou] (SCUT), Sun Yat-Sen University [Guangzhou] (SYSU), Universidad de Extremadura - University of Extremadura (UEX), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Laboratoire Jean Kuntzmann (LJK), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Spatial resolution, Hyperspectral imaging, Task analysis, Image reconstruction, Training, General Earth and Planetary Sciences, Convolutional neural networks, Pansharpening, Electrical and Electronic Engineering, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Traditional hyperspectral (HS) pansharpening aims at fusing a HS image with its panchromatic (PAN) counterpart, to bring the spatial resolution of the HS image to that of the PAN image. However, in many practical applications, arbitrary resolution HS (ARHS) pansharpening is required, where the HS and PAN images need to be integrated to generate a pansharpened HS image with arbitrary resolution (usually higher than that of the PAN image). Such an innovative task brings forth new challenges for the pansharpening technique, mainly including how to reconstruct HS images beyond the training scale and how to guarantee spectral fidelity at any spatial resolutions. To tackle the challenges, we present a novel convolutional neural network (CNN)-based method for ARHS pansharpening called ARHS-CNN. It is based on a two-step relay optimization process, which is associated with a multilevel enhancement subnetwork and a rescaling subnetwork. With a careful design following the thread, our ARHS-CNN is able to pansharpen HS images to any spatial resolutions using just a single CNN model trained on a limited number of scales while meantime to keep spectral fidelity at those resolutions, which wins an obvious advantage over traditional pansharpening methods. Experimental results obtained on several datasets verify the excellent performance of our ARHS-CNN method.
Published: 2022

19. ArbRPN: A Bidirectional Recurrent Pansharpening Network for Multispectral Images With Arbitrary Numbers of Bands

Author: Xiaomin Yang, Zhibing Lai, Gemine Vivone, Gwanggil Jeon, Jocelyn Chanussot, Lihui Chen, Sichuan University [Chengdu] (SCU), Institute of Methodologies for Environmental Analysis of the National Research Council (IMAA), National Research Council of Italy | Consiglio Nazionale delle Ricerche (CNR), Incheon National University, Chinese Academy of Sciences [Beijing] (CAS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: business.industry, Computer science, Multispectral image, pansharpening, Pattern recognition, multispec- tral (MS) images, image fusion, remote sensing, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Deep learning (DL), recurrent neural networks (RNNs), General Earth and Planetary Sciences, Artificial intelligence, Electrical and Electronic Engineering, business
Abstract: International audience; Although the performance of pansharpening has been significantly improved by advanced deep-learning (DL) technologies in recent years, most DL-based methods fail to process multispectral (MS) images with arbitrary numbers of bands by a single model. Consequently, it is inevitable to train separate models for MS images with different numbers of bands, which is time- and storage-consuming as well as inefficient in practice. To tackle the above problem, we propose a bidirectional recurrent pansharpening network (named ArbRPN) for MS images with arbitrary numbers of bands. Our ArbRPN can dynamically reconstruct high-resolution (HR) MS images with different numbers of bands by adaptively changing the number of recurrence to the number of bands of the low-resolution (LR) MS images. Leveraging on the ability of the ArbRPN to process MS images with any number of bands, one can even customize the bands to be pansharpened. Moreover, to achieve superior performance, spectral discrepancy and dependence are considered in the ArbRPN. Details from the panchromatic (PAN) image are adaptively injected into the fused product according to the captured spectral dependence. Furthermore, training strategies of existing DL-based pansharpening methods can only group MS images with a constant number of bands into mini-batches. Therefore, we present a mask-based training method (called mask-training) to solve this problem. Benefiting from the mask-training, our ArbRPN can achieve superior performance and robustness during pansharpening. Extensive experiments show the superior performance of our ArbRPN with respect to the state-of-the-art (SOTA) methods applied to MS images with different numbers of bands. The code of our ArbRPN is available on https://github.com/Lihui-Chen/ArbRPN.git.
Published: 2022

20. Tethered Drone-Based Airborne Wind Energy System Launching and Retrieving

Author: Jonathan Dumon, Ahmad Hably, Alexandre Sarazin, Amaury Nègre, Audrey Schanen, Nacim Meslem, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), GIPSA-Services (GIPSA-Services), GIPSA - Modelling and Optimal Decision for Uncertain Systems (GIPSA-MODUS), and GIPSA Pôle Automatique et Diagnostic (GIPSA-PAD)
Subjects: 0209 industrial biotechnology, Wind power, Angle of attack, business.industry, 020209 energy, Applied Mathematics, Aerospace Engineering, 02 engineering and technology, Rudder, Drone, Takeoff and landing, [SPI]Engineering Sciences [physics], Ground station, 020901 industrial engineering & automation, Space and Planetary Science, Control and Systems Engineering, 0202 electrical engineering, electronic engineering, information engineering, Environmental science, Electrical and Electronic Engineering, business, ComputingMilieux_MISCELLANEOUS, Marine engineering
Abstract: International audience
Published: 2021

21. Imagerie Hyperspectrale Et Zones Urbaines: Résultats Du Projet HYEP

Author: Weber, Christiane, Briottet, Xavier, Houet, Thomas, Gadal, Sébastien, Aguejdad, Rahim, Deville, Y., Mura, Mauro Dalla, Mallet, Clément, Le Bris, Arnaud, Karoui, Moussa Sofiane, Benhalouche, Fatima Zohra, Djerriri, Khelifa, Fabre, Sophie, Aval, Josselin, Territoires, Environnement, Télédétection et Information Spatiale (UMR TETIS), Centre de Coopération Internationale en Recherche Agronomique pour le Développement (Cirad)-AgroParisTech-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE), ONERA / DOTA, Université de Toulouse [Toulouse], ONERA-PRES Université de Toulouse, Littoral, Environnement, Télédétection, Géomatique (LETG - Rennes ), Université de Brest (UBO)-Université de Rennes 2 (UR2)-Nantes Université (Nantes Univ)-Littoral, Environnement, Télédétection, Géomatique UMR 6554 (LETG), Université de Brest (UBO)-Université de Rennes 2 (UR2)-Centre National de la Recherche Scientifique (CNRS)-Institut de Géographie et d'Aménagement Régional de l'Université de Nantes (Nantes Univ - IGARUN), Nantes Université - pôle Humanités, Nantes Université (Nantes Univ)-Nantes Université (Nantes Univ)-Nantes Université - pôle Humanités, Nantes Université (Nantes Univ)-Nantes Université (Nantes Univ)-Centre National de la Recherche Scientifique (CNRS)-Institut de Géographie et d'Aménagement Régional de l'Université de Nantes (Nantes Univ - IGARUN), Nantes Université (Nantes Univ), Aix Marseille Université (AMU), Études des Structures, des Processus d’Adaptation et des Changements de l’Espace (ESPACE), Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Avignon Université (AU)-Aix Marseille Université (AMU)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA), North-Eastern Federal University, Institut de recherche en astrophysique et planétologie (IRAP), Université Toulouse III - Paul Sabatier (UT3), Université de Toulouse (UT)-Université de Toulouse (UT)-Institut national des sciences de l'Univers (INSU - CNRS)-Observatoire Midi-Pyrénées (OMP), Institut de Recherche pour le Développement (IRD)-Université Toulouse III - Paul Sabatier (UT3), Université de Toulouse (UT)-Université de Toulouse (UT)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National d'Études Spatiales [Toulouse] (CNES)-Centre National de la Recherche Scientifique (CNRS)-Météo-France -Institut de Recherche pour le Développement (IRD)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National d'Études Spatiales [Toulouse] (CNES)-Centre National de la Recherche Scientifique (CNRS)-Météo-France -Centre National de la Recherche Scientifique (CNRS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Institut Universitaire de France (IUF), Ministère de l'Education nationale, de l’Enseignement supérieur et de la Recherche (M.E.N.E.S.R.), Laboratoire sciences et technologies de l'information géographique (LaSTIG), Ecole des Ingénieurs de la Ville de Paris (EIVP)-École nationale des sciences géographiques (ENSG), Institut National de l'Information Géographique et Forestière [IGN] (IGN)-Université Gustave Eiffel-Institut National de l'Information Géographique et Forestière [IGN] (IGN)-Université Gustave Eiffel, Université des sciences et de la Technologie d'Oran Mohamed Boudiaf [Oran] (USTO MB), Agence Spatiale Algérienne = Algerian Space Agency (ASAL), Centre National des Techniques Spatiales, Light Scatter Learning (LABISEN-LSL), Laboratoire ISEN (L@BISEN), Institut supérieur de l'électronique et du numérique (ISEN)-YNCREA OUEST (YO)-Institut supérieur de l'électronique et du numérique (ISEN)-YNCREA OUEST (YO), and ANR-14-CE22-0016,HYEP,Imagerie hyperspectrale pour la planification urbaine environnementale(2014)
Subjects: surfaces imperméables, fusion, impervious surface, panneaux photovoltaïques, Toulouse, [SHS]Humanities and Social Sciences, classification, vegetation, Hyperspectral imagery, [SDE]Environmental Sciences, Image Hyperspectrale, [INFO]Computer Science [cs], végétation, solar panels
Abstract: International audience; he HYEP project (ANR HYEP 14-CE22-0016-01 Hyperspectral imagery for Environmental urban Planning - Mobility and Urban Systems Programme - 2014) confirmed the interest of a global approach to the urban environment by remote sensing and in particular by using hyperspectral imaging (HI). The interest of hyperspectral images lies in the range of information provided over wavelengths from 0.4 to 4 μm; they thus provide access to spectral quantities of interest and to chemical or biophysical parameters of the surface. HYEP's objective was to specify this and to propose a panel of methods and treatments taking into account the characteristics of other existing sensors in order to compare their performance. The developments carried out were applied and evaluated on hyperspectral airborne images acquired in Toulouse and Kaunas (Lithuania), also used to synthesize space systems: Sentinel-2, Hypxim/Biodiversity and Pleiades. Among the locks identified were those related to improving the spatial capabilities of the sensors and spatial scale changes, which were partially overcome by fusion and sharpening approaches, which proved to be successful. After a description of our hyperspectral data set acquired over Toulouse, an analysis is conducted on several existing and accessible spectral databases. Then, the chosen methods are presented. They include extraction, fusion and classification methods, which are then applied on our dataset synthetized at different spatial resolution to evaluate the benefits and the complementarity of hyperspectral imagery in comparison with other traditional sensors. Some specific applications are investigated of interest for urban planners: impervious soil map, vegetation species cartography and detection of solar panels. Finally, discussion and perspectives are presented.; Le projet HYEP (ANR HYEP 14-CE22-0016-01 HYperspectral imagery for Environmental urban Planning - Programme Mobilité et systèmes urbains - 2014) a permis de confirmer l’intérêt d’une approche globale du milieu urbain par télédétection et notamment en utilisant l’imagerie hyperspectrale (IH). L’intérêt des images hyperspectrales réside dans la gamme d’informations fournies sur des longueurs d’onde de 0,4 à 4 μm ; elles fournissent ainsi un accès aux grandeurs spectrales d'intérêt et aux paramètres chimiques ou biophysiques de la surface. HYEP avait pour objectif de préciser ceci et de proposer un panel de méthodes et de traitements en tenant compte des caractéristiques des autres capteurs existants afin d’en comparer les performances. Les développements réalisés ont été appliqués et évalués sur des images aéroportées hyperspectrales acquises sur Toulouse (France) et Kaunas (Lituania), utilisées également pour synthétiser des systèmes spatiaux : Sentinel-2, Hypxim/Biodiversity et Pleiades. Parmi les verrous identifiés, ceux liés à l'amélioration des capacités spatiales des capteurs et aux changements d’échelle spatiale ont été en partie levés par des approches de fusion, qui se sont avérées concluantes. Des méthodes d’extraction, de fusion ou de classification ont été appliquées à différentes résolutions spatiales afin de qualifier les gains et la complémentarité de l’IH par rapport à d’autres capteurs. Ces évaluations ont ainsi porté sur des « objets » d’intérêt pour les gestionnaires et aménageurs du milieu urbain : les surfaces imperméables, la végétation ou encore la détection de panneaux solaires.
Published: 2022

22. DUNE: Deep UNcertainty Estimation for tracked visual features

Author: Lillo, Katia, de Maio, Andrea, Lacroix, Simon, Nègre, Amaury, Rombaut, Michèle, Marchand, Nicolas, Vercier, Vercier, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Équipe Robotique et InteractionS (LAAS-RIS), Laboratoire d'analyse et d'architecture des systèmes (LAAS), Université Toulouse Capitole (UT Capitole), Université de Toulouse (UT)-Université de Toulouse (UT)-Institut National des Sciences Appliquées - Toulouse (INSA Toulouse), Institut National des Sciences Appliquées (INSA)-Université de Toulouse (UT)-Institut National des Sciences Appliquées (INSA)-Université Toulouse - Jean Jaurès (UT2J), Université de Toulouse (UT)-Université Toulouse III - Paul Sabatier (UT3), Université de Toulouse (UT)-Centre National de la Recherche Scientifique (CNRS)-Institut National Polytechnique (Toulouse) (Toulouse INP), Université de Toulouse (UT)-Université Toulouse Capitole (UT Capitole), Université de Toulouse (UT), GIPSA-Services (GIPSA-Services), Thales avionics, and Marchand, Nicolas
Subjects: [INFO.INFO-RB] Computer Science [cs]/Robotics [cs.RO], [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO], [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, [SPI.SIGNAL] Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Uncertainty estimation of visual feature is essential for vision-based systems, such as visual navigation. We show that errors inherent to visual tracking, in particular using KLT tracker, can be learned using a probabilistic loss function to estimate the covariance matrix on each tracked feature position. The proposed system is trained and evaluated on synthetic data, as well as on real data, highlighting good results in comparison to the state of the art. The benefits of the tracking uncertainty estimates are illustrated for visual motion estimation.
Published: 2022

23. Minimalistic in-flight odometry based on two optic flow sensors along a bouncing trajectory

Author: Bergantin, Lucia, Coquet, Charles, Nègre, Amaury, Raharijaona, Thibaut, Marchand, Nicolas, Ruffier, Franck, Aix Marseille Université (AMU), Institut des Sciences du Mouvement Etienne Jules Marey (ISM), Aix Marseille Université (AMU)-Centre National de la Recherche Scientifique (CNRS), GIPSA-Services (GIPSA-Services), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA), HESAM Université - Communauté d'universités et d'établissements Hautes écoles Sorbonne Arts et métiers université (HESAM), GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), and Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab)
Subjects: [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO], [SPI.AUTO]Engineering Sciences [physics]/Automatic
Abstract: International audience; Estimating the distance traveled while navigating in a GPS-deprived environment is key for aerial robotic applications. For drones, this issue is often coupled with weight and computational power constraints, from which stems the importance of minimalistic equipment. In this study, we present a visual odometry strategy based solely on two optic flow magnitudes perceived by two optic flow sensors oriented at ±30° on either side of a drone's vertical axis. As results, (i) we measured the local optic flow divergence and the local translational optic flow respectively as the subtraction and the sum of the two optic flow magnitudes perceived (ii) we validated experimentally the visual odometer on a hexarotor oscillating upand-down while following a 50m-long circular trajectory under three illuminance conditions (117lux, 814lux and 1518lux). The measured optic flow divergence was used to estimate the flight height by means of an Extended Kalman Filter. The estimated flight height scaled the measured translational optic flow, which was integrated to perform minimalistic visual odometry.
Published: 2022

24. Sub-Pixel Optical Satellite Image Registration For Ground Deformation Using Deep Learning

Author: Tristan Montagnon, James Hollingsworth, Erwan Pathier, Mathilde Marchandon, Mauro Dalla Mura, Sophie Giffard-Roisin, Institut des Sciences de la Terre (ISTerre), Institut national des sciences de l'Univers (INSU - CNRS)-Institut de recherche pour le développement [IRD] : UR219-Université Savoie Mont Blanc (USMB [Université de Savoie] [Université de Chambéry])-Centre National de la Recherche Scientifique (CNRS)-Université Gustave Eiffel-Université Grenoble Alpes (UGA), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Tokyo Institute of Technology [Tokyo] (TITECH), Institut Universitaire de France (IUF), Ministère de l'Education nationale, de l’Enseignement supérieur et de la Recherche (M.E.N.E.S.R.), and projet IRGA Call 2021 'Initiatives de Recherche à Grenoble Alpes'
Subjects: image registration, optical image correlation, [STAT.ML]Statistics [stat]/Machine Learning [stat.ML], [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, ACM: I.: Computing Methodologies/I.4: IMAGE PROCESSING AND COMPUTER VISION, geodesy, deep learning, ACM: F.: Theory of Computation/F.1: COMPUTATION BY ABSTRACT DEVICES/F.1.1: Models of Computation/F.1.1.4: Self-modifying machines (e.g., neural networks), [SDU.STU.GM]Sciences of the Universe [physics]/Earth Sciences/Geomorphology, satellite imagery, ACM: I.: Computing Methodologies/I.4: IMAGE PROCESSING AND COMPUTER VISION/I.4.3: Enhancement/I.4.3.3: Registration
Abstract: International audience; Precise estimation of ground displacement maps at regional scales from optical satellite imaging is fundamental for the comprehension of natural disasters, such as earthquakes. Current methods make use of correlation techniques between two acquisitions in order to retrieve a fractional pixel shift. Yet, differences in local lighting conditions between the two acquisitions can lead to high image differences which will bias the estimation of the displacement, and data-driven methods could have the ability to overcome these errors. From the generation of a realistic simulated database based on Landsat-8 satellite pairs of images with added simulated shifts, we developed a Convolutional Neural Network (CNN) able to retrieve a sub-pixel displacement.
Published: 2022

25. Process Oriented evaluation of the oversea AROME configuration: focus on the representation of cloud organisation

Author: Florent Beucher, Fleur Couvreux, Dominique Bouniol, Ghislain Faure, Florence Favot, Thibaut Dauhut, Alex Ayet, Centre national de recherches météorologiques (CNRM), Institut national des sciences de l'Univers (INSU - CNRS)-Observatoire Midi-Pyrénées (OMP), Institut de Recherche pour le Développement (IRD)-Université Toulouse III - Paul Sabatier (UT3), Université de Toulouse (UT)-Université de Toulouse (UT)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National d'Études Spatiales [Toulouse] (CNES)-Centre National de la Recherche Scientifique (CNRS)-Météo-France -Institut de Recherche pour le Développement (IRD)-Université Toulouse III - Paul Sabatier (UT3), Université de Toulouse (UT)-Université de Toulouse (UT)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National d'Études Spatiales [Toulouse] (CNES)-Centre National de la Recherche Scientifique (CNRS)-Météo-France -Centre National de la Recherche Scientifique (CNRS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Climat, Environnement, Couplages et Incertitudes [Toulouse] (CECI), and Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique (CERFACS)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National de la Recherche Scientifique (CNRS)
Subjects: [SDU.OCEAN]Sciences of the Universe [physics]/Ocean, Atmosphere, Atmospheric Science, evaluation, Clouds < 3. Physical phenomenon Boundary layer < 4. Geophysical sphere Tropics < 5. Geographic/climatic zone Dynamic/Processes < 1, [SDU]Sciences of the Universe [physics], Tropics < 5. Geographic/climatic zone, cloud resolving model, shallow convection, North Atlantic Trades, Boundary layer < 4. Geophysical sphere, cloud -organisation, Clouds < 3. Physical phenomenon, Dynamic/Processes < 1
Abstract: International audience; This study evaluates the ability of the French convection-permitting model AROME-OM to represent shallow cumulus and their main organisations forboreal winter conditions in the North Atlantic trades. It uses a set of three winter seasons (January–February, 2018–2020) of high-resolution (1.3 and 2.5 km) simulations over the Caribbean domain (9.7–22.9◦ N, 75.3–51.7◦ W). The model is assessed against soundings at Grantley Adams Airport and remote-sensing observations at a site located on the east coast of Barbados which is representative of downwind trade regimes. The thermodynamic environment of the model fits the observations overall but the boundary layer is slightly too deep, resulting in a weak cold and dry bias. Both model and observations clearly exhibit (a) a double peak of cloud fraction, a first peak near the cloud base and a second one near the cloud top and (b) a larger variance in cloudiness near the top of the deepest cumuli, at around 2 km, with a higher sensitivity to the environment. We then take advantage of the EUREC4 A field campaign which took place in January–February 2020 to assess the ability of the model to reproduce the four main mesoscale patterns and to characterize the air masses in which they develop. All the observations confirm the capacity of the model to predict the different mesoscale organizations and their associated environment.
Published: 2022

26. Onboard dynamic RGB‐D simultaneous localization and mapping for mobile robot navigation

Author: Bruce Canovas, Amaury Nègre, Michèle Rombaut, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), and GIPSA-Services (GIPSA-Services)
Subjects: dynamic environment, TK7800-8360, General Computer Science, Computer science, TK5101-6720, 02 engineering and technology, Simultaneous localization and mapping, 01 natural sciences, rgb‐d, 0202 electrical engineering, electronic engineering, information engineering, [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO], Computer vision, Electrical and Electronic Engineering, navigation, Dense SLAM, business.industry, 010401 analytical chemistry, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], RGB-D, 020207 software engineering, Mobile robot navigation, 0104 chemical sciences, Electronic, Optical and Magnetic Materials, mobile robotics, [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV], Telecommunication, Artificial intelligence, Electronics, business
Abstract: International audience; Although the actual visual simultaneous localization and mapping (SLAM) algorithms provide highly accurate tracking and mapping, most algorithms are too heavy to run live on embedded devices. In addition, the maps they produce are often unsuitable for path planning. To mitigate these issues, we propose a completely closed-loop online dense RGB-D SLAM algorithm targeting autonomous indoor mobile robot navigation tasks. The proposed algorithm runs live on an NVIDIA Jetson board embedded on a two-wheel differential-drive robot. It exhibits lightweight three-dimensional mapping, room-scale consistency, accurate pose tracking, and robustness to moving objects. Further, we introduce a navigation strategy based on the proposed algorithm. Experimental results demonstrate the robustness of the proposed SLAM algorithm, its computational efficiency, and its benefits for on-the-fly navigation while mapping.
Published: 2021

27. An attention-fused network for semantic segmentation of very-high-resolution remote sensing imagery

Author: Xiuping Jia, Shanshan Li, Xuan Yang, Jocelyn Chanussot, Baipeng Li, Zhengchao Chen, Bing Zhang, Pan Chen, Chinese Academy of Sciences [Beijing] (CAS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Australian Defence Force Academy [Canberra] (ADFA), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: FOS: Computer and information sciences, 010504 meteorology & atmospheric sciences, Computer science, Computer Vision and Pattern Recognition (cs.CV), Computer Science - Computer Vision and Pattern Recognition, 0211 other engineering and technologies, Convolutional neural network, 02 engineering and technology, Very-high-resolution imagery, 01 natural sciences, Segmentation, Computers in Earth Sciences, Engineering (miscellaneous), 021101 geological & geomatics engineering, 0105 earth and related environmental sciences, Remote sensing, Block (data storage), Attention-fused network, business.industry, Deep learning, ISPRS, Sensor fusion, Semantic segmentation, Atomic and Molecular Physics, and Optics, Computer Science Applications, Artificial intelligence, F1 score, business, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Encoder, Multipath propagation
Abstract: Semantic segmentation is an essential part of deep learning. In recent years, with the development of remote sensing big data, semantic segmentation has been increasingly used in remote sensing. Deep convolutional neural networks (DCNNs) face the challenge of feature fusion: very-high-resolution remote sensing image multisource data fusion can increase the network's learnable information, which is conducive to correctly classifying target objects by DCNNs; simultaneously, the fusion of high-level abstract features and low-level spatial features can improve the classification accuracy at the border between target objects. In this paper, we propose a multipath encoder structure to extract features of multipath inputs, a multipath attention-fused block module to fuse multipath features, and a refinement attention-fused block module to fuse high-level abstract features and low-level spatial features. Furthermore, we propose a novel convolutional neural network architecture, named attention-fused network (AFNet). Based on our AFNet, we achieve state-of-the-art performance with an overall accuracy of 91.7% and a mean F1 score of 90.96% on the ISPRS Vaihingen 2D dataset and an overall accuracy of 92.1% and a mean F1 score of 93.44% on the ISPRS Potsdam 2D dataset., Comment: 35 pages. Published by ISPRS Journal of Photogrammetry and Remote Sensing
Published: 2021

28. Hidden Semi-Markov Models to Segment Reading Phases from Eye Movements

Author: Brice Olivier, Anne Guérin-Dugué, Jean-Baptiste Durand, Modèles statistiques bayésiens et des valeurs extrêmes pour données structurées et de grande dimension (STATIFY), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA), GIPSA - Vision and Brain Signal Processing (GIPSA-VIBS), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and ANR-11-LABX-0025,PERSYVAL-lab,Systemes et Algorithmes Pervasifs au confluent des mondes physique et numérique(2011)
Subjects: hidden semi-Markov chains, Ophthalmology, Eye movement, scanpath, reading, [MATH.MATH-ST]Mathematics [math]/Statistics [math.ST], [SCCO.NEUR]Cognitive science/Neuroscience, segmentation, eye tracking, individual differences, Sensory Systems
Abstract: International audience; Our objective is to analyze scanpaths acquired through participants achieving a reading task aiming at answering a binary question: Is the text related or not to some given target topic? We propose a data-driven method based on hidden semi-Markov chains to segment scanpaths into phases deduced from the model states, which are shown to represent different cognitive strategies: normal reading, fast reading, information search, and slow confirmation. These phases were confirmed using different external covariates, among which semantic information extracted from texts. Analyses highlighted some strong preference of specific participants for specific strategies and more globally, large individual variability in eye-movement characteristics, as accounted for by random effects. As a perspective, the possibility of improving reading models by accounting for possible heterogeneity sources during reading is discussed.
Published: 2022

29. Multi Spectral-Spatial Gabor Feature Fusion Based On End-To-End Deep Learning For Hyperspectral Image Classification

Author: Refka Hanachi, Akrem Sellami, Imed Riadh Farah, Mauro Dalla Mura, Laboratoire de recherche en Génie Logiciel, Applications distribuées, Systèmes décisionnels et Imagerie intelligente [Manouba] (RIADI), École Nationale des Sciences de l'Informatique [Manouba] (ENSI), Université de la Manouba [Tunisie] (UMA)-Université de la Manouba [Tunisie] (UMA), Laboratoire Lorrain de Recherche en Informatique et ses Applications (LORIA), Institut National de Recherche en Informatique et en Automatique (Inria)-Université de Lorraine (UL)-Centre National de la Recherche Scientifique (CNRS), Département lmage et Traitement Information (IMT Atlantique - ITI), IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and Université Grenoble Alpes (UGA)
Subjects: Signal processing, [INFO.INFO-LG]Computer Science [cs]/Machine Learning [cs.LG], Feature extraction, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], Deep learning, Convolutional neural networks, Representation learning, [INFO.INFO-AI]Computer Science [cs]/Artificial Intelligence [cs.AI]
Abstract: International audience; The use of low-level spatial information, in addition to the rich spectral information of Hyperspectral Image (HSI) with the potential of deep learning-based methods, has provided high performance in improving the HSI classification. With this regard, Gabor Filtering has been successfully applied for HSI pixels analysis due to its ability in extracting representative spatial features. Therefore, a novel methodology based on Gabor filtering and multi-view Convolutional Neural Networks (multi-view CNNs) was proposed in this paper. Firstly, Gabor texture features at different scales and orientations performed on the first three principal components of HSI were extracted. Then, spectral and spatial features are introduced separately through spectral, and spatial CNN, respectively. Afterward, their learned feature maps are combined and fed into our spectral-spatial CNN (SSCNN) to learn a fused multi-view representation. Experiments applied on two real HSI datasets, including Indian Pines, and Salinas yield competitive classification performances compared to other state-of-the-art deep learning methods.
Published: 2022

30. Using trajectory oscillation timing improves in-flight odometry based solely on optic flows

Author: Lucia Bergantin, Charles Coquet, Amaury Nègre, Raharijoana Thibault Raharijoana, Nicolas Marchand, Franck Ruffier, Institut des Sciences Moléculaires de Marseille (ISM2), Aix Marseille Université (AMU)-École Centrale de Marseille (ECM)-Institut de Chimie du CNRS (INC)-Centre National de la Recherche Scientifique (CNRS), GIPSA-Services (GIPSA-Services), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA), Laboratoire de Conception Fabrication Commande (LCFC), Université de Lorraine (UL)-Arts et Métiers Sciences et Technologies, HESAM Université - Communauté d'universités et d'établissements Hautes écoles Sorbonne Arts et métiers université (HESAM)-HESAM Université - Communauté d'universités et d'établissements Hautes écoles Sorbonne Arts et métiers université (HESAM), GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Institut des Sciences du Mouvement Etienne Jules Marey (ISM), and Aix Marseille Université (AMU)-Centre National de la Recherche Scientifique (CNRS)
Subjects: [INFO.INFO-AU]Computer Science [cs]/Automatic Control Engineering, [INFO.INFO-RB]Computer Science [cs]/Robotics [cs.RO], [SPI.AUTO]Engineering Sciences [physics]/Automatic
Abstract: International audience; Traveled distance estimation is a common problem for robotic applications taking place in unknown environments where GPS is not available. In drones, the presence of weight and computational power constraints leads to the importance of developing odometry strategies based on minimilastic equipment. In this study, we imposed upon a hexarotor to perform up-and-down oscillatory motions while flying forward to test a self-scaled scheme of a visual odometer for the first time. For the odometry, the downward translational optic flow was scaled by the current visually estimated flight height and then mathematically integrated to evaluate the total distance traveled. The self-oscillatory trajectory generated successions of contraction and expansion in the optic flow vector field, which allowed to estimate the flight height of the hexarotor by means of an Extended Kalman Filter. We present three strategies based on sensor fusion that rely on no, precise or rough prior knowledge of the optic flow variations imposed by the sinusoidal trajectory. The rough prior knowledge strategy uses solely the timing of the variations of the optic flow. Tests were performed in a flying arena, where the hexarotor followed a circular trajectory while oscillating up-and-down over about 50m under illuminances of 117lux and 1518lux.
Published: 2022

31. A New Approach to Measure Vocal Tract Resonances. Application to a Modern Singing Study

Author: Maison, Timothée, Silva, Fabrice, Guillemain, Philippe, Henrich Bernardoni, Nathalie, Laboratoire de Mécanique et d'Acoustique [Marseille] (LMA ), Aix Marseille Université (AMU)-École Centrale de Marseille (ECM)-Centre National de la Recherche Scientifique (CNRS), GIPSA - Analyse et Modification de l'homme en mouvement : biomécanique, cognition, vocologie (GIPSA-MOVE), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and Université Grenoble Alpes (UGA)
Subjects: [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, [PHYS.PHYS.PHYS-INS-DET]Physics [physics]/Physics [physics]/Instrumentation and Detectors [physics.ins-det], [PHYS.MECA.ACOU]Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph]
Abstract: International audience; Measuring vocal-tract acoustics is one key issue for singing voice characterization. We aim at developing a non-invasive experimental tool that would assess the singer's vocal-tract acoustical behavior during singing. Grounded on previous impedance measurements, a broadband external acoustic source (chirp signal) loads vocal tract at the lips while phonating. Both chirp and voice signals are recorded by a microphone. An open-mouth condition measurement calibrated by a closed-mouth condition provides vocal-tract frequency response, from which resonance frequencies and quality factors can be estimated. A challenging aspect is to separate voice from chirp signal. We propose a new method for signal separation and test it numerically. Its applicability to real singing is then explored on vowels sung by a modern-style singer.First step consists in estimating voice fundamental frequency on recorded voice+chirp signal. Secondly, an informed high-order chirplet analysis is performed on this signal to eliminate voice, recovering only modulated chirp. Numerical simulations, based on synthesizing vowel [a] sung at 450Hz while adding a 3dB-lower chirp, show that the method is accurate enough for resonance frequencies (resp. quality factors) not to deviate by less than 3 cents (resp. 5%).The method is applied on recordings of a modern singer and singing teacher. The subject was asked to sing vowel [a] with different voice qualities and for several pitches (F3, G3, A3). Measurement was done during a one-second sustained phonation (voice condition) and during a following closed-glottis condition with instruction to keep vocal tract in same position (silence condition). Voice condition forces to apply voice-chirp separation method, on contrary to silence condition. Results show that the closed-glottis instruction causes additional vocal tract adjustments, even for a trained singer. The differences in resonances between registers are explored as a function of pitch and voice quality.
Published: 2022

32. Machine Learning Approaches to Automatically Detect Glacier Snow Lines on Multi-Spectral Satellite Images

Author: Colin Prieur, Antoine Rabatel, Jean-Baptiste Thomas, Ivar Farup, Jocelyn Chanussot, Norwegian University of Science and Technology [Gjøvik] (NTNU), Norwegian University of Science and Technology (NTNU), Data Science Experts (DSE), Institut des Géosciences de l’Environnement (IGE), Institut de Recherche pour le Développement (IRD)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: active contours, satellite remote sensing, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, SVM, General Earth and Planetary Sciences, glacier snow line, neural networks, multi-spectral images, random forest
Abstract: International audience; Documenting the inter-annual variability and the long-term trend of the glacier snow line altitude is highly relevant to document the evolution of glacier mass changes. Automatically identifying the snow line on glaciers is challenging; recent developments in machine learning approaches show promise to tackle this issue. This manuscript presents a proof of concept of machine learning approaches applied to multi-spectral images to detect the snow line and quantify its average altitude. The tested approaches include the combination of different image processing and classification methods, and takes into account cast shadows. The efficiency of these approaches is evaluated on mountain glaciers in the European Alps by comparing the results with manually annotated data. Solutions provided by the different approaches are robust when compared to the ground truth’s snow lines, with a Pearson’s correlation ranging from 79% to 96% depending on the method. However, the tested approaches may fail when snow lines are not continuous or exhibit a strong change of elevation. The major advantage over the state of the art is that the proposed approach does not require one calibration per glacier.
Published: 2022
Full Text: View/download PDF

33. Element-Wise Feature Relation Learning Network for Cross-Spectral Image Patch Matching

Author: Dou Quan, Shuang Wang, Jocelyn Chanussot, Xuefeng Liang, Licheng Jiao, Ning Huyan, Biao Hou, Wang Ruojing, Xidian University, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Matching (statistics), Aggregates, Computer Networks and Communications, Computer science, Generalization, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Convolutional neural network, Image (mathematics), Discriminative model, Artificial Intelligence, Training, Measurement, Learning systems, Image matching, business.industry, Pattern recognition, Computer Science Applications, Feature (computer vision), Task analysis, Feature extraction, Artificial intelligence, business, Focus (optics), Feature learning, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Software
Abstract: International audience; Recently, the majority of successful matching approaches are based on convolutional neural networks, which focus on learning the invariant and discriminative features for individual image patches based on image content. However, the image patch matching task is essentially to predict the matching relationship of patch pairs, that is, matching (similar) or non-matching (dissimilar). Therefore, we consider that the feature relation (FR) learning is more important than individual feature learning for image patch matching problem. Motivated by this, we propose an element-wise FR learning network for image patch matching, which transforms the image patch matching task into an image relationship-based pattern classification problem and dramatically improves generalization performances on image matching. Meanwhile, the proposed element-wise learning methods encourage full interaction between feature information and can naturally learn FR. Moreover, we propose to aggregate FR from multilevels, which integrates the multiscale FR for more precise matching. Experimental results demonstrate that our proposal achieves superior performances on cross-spectral image patch matching and single spectral image patch matching, and good generalization on image patch retrieval.
Published: 2022

34. ImSPOC: A Novel Compact Hyperspectral Camera for the Monitoring of Atmospheric Gases

Author: Gousset, Silvère, Picone, Daniele, Le Coarer, Etienne, Rodrigo, Juana, Voisin, D., Croizé, Laurence, Ferrec, Yann, Dalla Mura, Mauro, Institut de Planétologie et d'Astrophysique de Grenoble (IPAG), Centre National d'Études Spatiales [Toulouse] (CNES)-Observatoire des Sciences de l'Univers de Grenoble (OSUG ), Institut national des sciences de l'Univers (INSU - CNRS)-Université Savoie Mont Blanc (USMB [Université de Savoie] [Université de Chambéry])-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE)-Université Grenoble Alpes (UGA)-Météo-France -Institut national des sciences de l'Univers (INSU - CNRS)-Université Savoie Mont Blanc (USMB [Université de Savoie] [Université de Chambéry])-Centre National de la Recherche Scientifique (CNRS)-Institut National de Recherche pour l’Agriculture, l’Alimentation et l’Environnement (INRAE)-Université Grenoble Alpes (UGA)-Météo-France, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Institut des Géosciences de l’Environnement (IGE), Institut de Recherche pour le Développement (IRD)-Institut national des sciences de l'Univers (INSU - CNRS)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), DOTA, ONERA, Université Paris Saclay [Palaiseau], ONERA-Université Paris-Saclay, Institut Universitaire de France (IUF), Ministère de l'Education nationale, de l’Enseignement supérieur et de la Recherche (M.E.N.E.S.R.), and Tokyo Institute of Technology [Tokyo] (TITECH)
Subjects: Accident prevention, [PHYS]Physics [physics], Hyperspectral imaging, Monitoring, SPECTROSCOPIE, Hyperspectral imaging sensor, Miniaturized sensor, [SPI]Engineering Sciences [physics], Gas detector, SURVEILLANCE, Fourier transform, Gas monitoring, IMAGERIE HYPERSPECTRALE, DETECTION GAZ, Spectroscopy, PREVENTION ACCIDENT
Abstract: International audience; There is an increasing industrial and scientific demand for large scale monitoring of gas concentration, e.g., to abide with constraints on pollutants for environmental safety. This paper aims to show the potential of the very recent ImSPOC technology for gas monitoring. We present the ImSPOC operating principle, some of the developed prototypes and two illustrative examples in the monitoring of methane and carbon dioxide in controlled environments.
Published: 2022

35. Decoupled Control of a Twin Hull-Based Unmanned Surface Vehicle Using a Linear Parameter Varying Approach

Author: Echrak Chnib, Olivier Sename, Francesco Ferrante, Eduardo S. Rodriguez Canales, Juan C. Cutipa Luque, Ferrante, Francesco, GIPSA - COntrol, PErception, Robots, navigation and Intelligent Computing (GIPSA-COPERNIC), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), GIPSA - Infinite Dimensional Dynamics (GIPSA-INFINITY), GIPSA Pôle Automatique et Diagnostic (GIPSA-PAD), and Universidad Nacional de San Agust ́ın de Arequipa
Subjects: Decoupled control, [INFO.INFO-AU]Computer Science [cs]/Automatic Control Engineering, ∞, Unmanned surface vehicle (USV), [INFO.INFO-AU] Computer Science [cs]/Automatic Control Engineering, robust control, ComputingMilieux_MISCELLANEOUS, Linear parameter varying (LPV)
Abstract: International audience
Published: 2022

36. Variable Subpixel Convolution Based Arbitrary-Resolution Hyperspectral Pansharpening

Author: Lin He, Jinhua Xie, Jun Li, Antonio Plaza, Jocelyn Chanussot, Jiawei Zhu, South China University of Technology [Guangzhou] (SCUT), China University of Geosciences [Wuhan] (CUG), Universidad de Extremadura - University of Extremadura (UEX), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Spatial resolution, Standards, Task analysis, General Earth and Planetary Sciences, Training, Pansharpening, Electrical and Electronic Engineering, Optimized production technology, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Convolution
Abstract: International audience; Standard hyperspectral (HS) pansharpening relies on fusion to enhance low-resolution HS (LRHS) images to the resolution of their matching panchromatic (PAN) images, whose practical implementation is normally under a stipulation of scale invariance of the model across the training phase and the pansharpening phase. By contrast, arbitrary resolution HS (ARHS) pansharpening seeks to pansharpen LRHS images to any user-customized resolutions. For such a new HS pansharpening task, it is not feasible to train and store convolution neural network (CNN) models for all possible candidate scales, which implies that the single model acquired from the training phase should be capable of being generalized to yield HS images with any resolutions in the pansharpening phase. To address the challenge, a novel variable subpixel convolution (VSPC)-based CNN (VSPC-CNN) method following our arbitrary upsampling CNN (AU-CNN) framework is developed for ARHS pansharpening. The VSPC-CNN method comprises a two-stage elevating thread. The first stage is to improve the spatial resolution of the input HS image to that of the PAN image through a prepansharpening module, and then, a VSPC-encapsulated arbitrary scale attention upsampling (ASAU) module is cascaded for arbitrary resolution adjustment. After training with given scales, it can be generalized to pansharpen HS image to arbitrary scales under the spatial patterns invariance across the training and pansharpening phases. Experimental results from several specific VSPC-CNNs on both simulated and real HS datasets show the superiority of the proposed method.
Published: 2022

37. Blind Hyperspectral Unmixing Based on Graph Total Variation Regularization

Author: Jocelyn T. Chi, Harlin Lee, Jocelyn Chanussot, Andrea L. Bertozzi, Lucas Drumetz, Jing Qin, Yifei Lou, Department of Mathematics, University of Kentucky, Lexington, Department of Electrical and Computer Engineering, Carnegie Mellon University, Pittsburgh, Department of Statistics, North Carolina State University, Raleigh, Equipe Observations Signal & Environnement (Lab-STICC_OSE), Laboratoire des sciences et techniques de l'information, de la communication et de la connaissance (Lab-STICC), Institut Mines-Télécom [Paris] (IMT)-IMT Atlantique Bretagne-Pays de la Loire (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-École Nationale d'Ingénieurs de Brest (ENIB)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Centre National de la Recherche Scientifique (CNRS)-Université Bretagne Loire (UBL)-Institut Mines-Télécom [Paris] (IMT)-IMT Atlantique Bretagne-Pays de la Loire (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-École Nationale d'Ingénieurs de Brest (ENIB)-École Nationale Supérieure de Techniques Avancées Bretagne (ENSTA Bretagne)-Université de Bretagne Sud (UBS)-Université de Brest (UBO)-Centre National de la Recherche Scientifique (CNRS)-Université Bretagne Loire (UBL), Département Mathematical and Electrical Engineering (IMT Atlantique - MEE), IMT Atlantique Bretagne-Pays de la Loire (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Department of Mathematical Sciences, University of Texas at Dallas, Richardson, Department of Mathematics [UCLA], University of California [Los Angeles] (UCLA), University of California-University of California, and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Computer science, 0211 other engineering and technologies, Nystrom method, 02 engineering and technology, blind hyperspectral unmixing, Geological & Geomatics Engineering, Blind hyperspectral unmixing, Regularization (mathematics), Spectral line, graph total variation, graph Laplacian, 0202 electrical engineering, electronic engineering, information engineering, Electrical and Electronic Engineering, Image resolution, 021101 geological & geomatics engineering, Pixel, business.industry, alternating direction method of multipliers, Hyperspectral imaging, Pattern recognition, Nyström method, Total variation denoising, Alternating direction method of multipliers, Geophysics, Geomatic Engineering, General Earth and Planetary Sciences, Graph (abstract data type), 020201 artificial intelligence & image processing, Artificial intelligence, Laplacian matrix, business, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Remote sensing data from hyperspectral cameras suffer from limited spatial resolution, in which a single pixel of a hyperspectral image may contain information from several materials in the field of view. Blind hyperspectral image unmixing is the process of identifying the pure spectra of individual materials (i.e., endmembers) and their proportions (i.e., abundances) at each pixel. In this paper, we propose a novel blind hyperspectral unmixing model based on the graph total variation (gTV) regularization, which can be solved efficiently by the alternating direction method of multipliers (ADMM). To further alleviate the computational cost, we apply the Nyström method to approximate a fully-connected graph by a small subset of sampled points. Furthermore, we adopt the Merriman-Bence-Osher (MBO) scheme to solve the gTV-involved subproblem in ADMM by decomposing a grayscale image into a bit-wise form. A variety of numerical experiments on synthetic and real hyperspectral images are conducted, showcasing the potential of the proposed method in terms of identification accuracy and computational efficiency.
Published: 2021

38. Hydrophone Array Optimization, Conception, and Validation for Localization of Acoustic Sources in Deep-Sea Mining

Author: Arthur Finez, Simon Bouley, Florent Fayet, Barbara Nicolas, Jerome Mars, Valentin Baron, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Imagerie Ultrasonore, Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé (CREATIS), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet - Saint-Étienne (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet - Saint-Étienne (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS), MicrodB, OSEAN S.A.S [Le Pradet], Osean, Nicolas, Barbara, Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet [Saint-Étienne] (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Hospices Civils de Lyon (HCL)-Université Jean Monnet [Saint-Étienne] (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS), Université Jean Monnet [Saint-Étienne] (UJM)-Hospices Civils de Lyon (HCL)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université de Lyon-Institut National des Sciences Appliquées (INSA)-Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Centre National de la Recherche Scientifique (CNRS)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Université Jean Monnet [Saint-Étienne] (UJM)-Hospices Civils de Lyon (HCL)-Institut National des Sciences Appliquées de Lyon (INSA Lyon), and Université de Lyon-Centre National de la Recherche Scientifique (CNRS)-Institut National de la Santé et de la Recherche Médicale (INSERM)
Subjects: Beamforming, Ground truth, source localization, Hydrophone, [INFO.INFO-TS] Computer Science [cs]/Signal and Image Processing, 010505 oceanography, Computer science, Mechanical Engineering, Acoustics, 020206 networking & telecommunications, Ocean Engineering, 02 engineering and technology, 01 natural sciences, deep-sea mining, Deep sea mining, underwater acoustics, Noise, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, 0202 electrical engineering, electronic engineering, information engineering, 14. Life underwater, Electrical and Electronic Engineering, Underwater, Underwater acoustics, Image resolution, 0105 earth and related environmental sciences
Abstract: International audience; As the mining of deep-sea natural deposits is becoming cost competitive compared to similar land-based mining, companies have started to dig into the seabeds to collect minerals. However, the acoustic contribution of this activity in the surrounding environment can be significant. To predict the impact of such noise, the starting point is to localize and quantify the sources that create it. In this study, a 3-D prototype acoustic array to perform this localization and quantification is designed, built, and deployed at sea for validation of its localization capacities. The design method performs a two-step study to define the array shape and select the hydrophone arrangement over it, under harsh constraints. Each step relies on two metrics to rank the candidates: the maximum sidelobe level, and the spatial resolution. These are computed on conventional beamforming maps for simulated sources that represent excavation machines on the ground. The shape is first determined to be the one that yields steady maximum sidelobe value levels over frequency. Second, the hydrophone arrangement that achieves the lowest maximum sidelobe level while limiting the spatial resolution is selected. This leads to a tip down conical array with 21 hydrophones, of about 3 m in height and diameter, and this is manufactured and used during an experimental campaign in the Mediterranean Sea. The experimental localization maps show strong agreement between the estimated source position and its ground truth. A more detailed comparison between simulated and real performances confirms accurate array conception and realization. Thus, this design procedure provides an efficient underwater acoustic array for monitoring deep-sea mining, the localization capacities of which are validated in a real-life setting.
Published: 2021

39. Cluster-Memory Augmented Deep Autoencoder via Optimal Transportation for Hyperspectral Anomaly Detection

Author: Ning Huyan, Xiangrong Zhang, Dou Quan, Jocelyn Chanussot, Licheng Jiao, Xidian University, GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Memory modules, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Image reconstruction, Task analysis, General Earth and Planetary Sciences, Prototypes, Detectors, Transportation, Anomaly detection, Electrical and Electronic Engineering
Abstract: International audience; Hyperspectral anomaly detection (AD) aims to detect objects significantly different from their surrounding background. Recently, many detectors based on autoencoder (AE) exhibited promising performances in hyperspectral AD tasks. However, the fundamental hypothesis of the AE-based detector that anomaly is more challenging to be reconstructed than background may not always be true in practice. We demonstrate that an AE could well reconstruct anomalies even without anomalies for training, because AE models mainly focus on the quality of sample reconstruction and do not care if the encoded features solely represent the background rather than anomalies. If more information is preserved than needed to reconstruct the background, the anomalies will be well reconstructed. This article proposes a cluster-memory augmented deep autoencoder via optimal transportation for hyperspectral anomaly detection (OTCMA) clustering for hyperspectral AD to solve this problem. The deep clustering method based on optimal transportation (OT) is proposed to enhance the features consistency of samples within the same categories and features discrimination of samples in different categories. The memory module stores the background’s consistent features, which are the cluster centers for each category background. We retrieve more consistent features from the memory module instead of reconstructing a sample utilizing its own encoded features. The network focuses more on consistent feature reconstruction by training AE with a memory module. This effectively restricts the reconstruction ability of AE and prevents reconstructing anomalies. Extensive experiments on the benchmark datasets demonstrate that our proposed OTCMA achieves state-of-the-art results. Besides, this article presents further discussions about the effectiveness of our proposed memory module and different criteria for better AD.
Published: 2022

40. Multigraph-Based Low-Rank Tensor Approximation for Hyperspectral Image Restoration

Author: Na Liu, Wei Li, Ran Tao, Qian Du, Jocelyn Chanussot, Beijing Institute of Technology (BIT), Mississippi State University [Mississippi], Chinese Academy of Sciences [Beijing] (CAS), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), and Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP )
Subjects: low-rank tensor, hyperspectral imagery, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, General Earth and Planetary Sciences, Graph signal processing, Electrical and Electronic Engineering, image restoration
Abstract: International audience; Low-rank-tensor-approximation (LRTA)-based hyperspectral image and hyperspectral imagery (HSI) restoration has drawn increasing attention. However, most of the methods construct a hidden low-rank tensor by utilizing the nonlocal self-similarity (NLSS) and global spectral correlation (GSC) inherited by HSIs. Although achieving state-of-the-art (SOTA) restoration performance, NLSS and GSC have limitations. NLSS is introduced from natural image denoising to remove spatially independent identically distributed (i.i.d.) Gaussian and impulse noise, while GSC, which is naturally possessed by HSIs, is adopted to maintain the spectral integrity and remove spectrally, i.i.d., degradations. Therefore, NLSS and GSC may not be successfully used for complex HSI restoration tasks, such as destriping, cloud removal, and recovery of atmospheric absorption bands.To solve the issue, borrowing the idea from manifold learning, the geometry information characterized by proximity relationship, is integrated with the LRTA to solve the above issue, named multigraph-based LRTA (MGLRTA). Different from most of the existing methods, the proposed MGLRTA directly models an HSI as a low-rank tensor and efficiently explores the extra proximity information on the defined graphs that are not only inherited by the low-rank constraints but also naturally possessedin HSIs. A well-posed iterative algorithm is designed to solve the restoration problem. Experimental results on different datasets that cover several severe degradation scenarios demonstrate that the proposed MGLRTA outperforms the SOTA HSI restoration methods.
Published: 2022

41. Le corps en images: Les nouvelles imageries pour la santé

Author: Blanc-Féraud, Laure, Caruyer, Emmanuel, Jutten, Christian, Liebgott, Herve, Morphologie et Images (MORPHEME), Inria Sophia Antipolis - Méditerranée (CRISAM), Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Institut de Biologie Valrose (IBV), Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Signal, Images et Systèmes (Laboratoire I3S - SIS), Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis (I3S), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Université Nice Sophia Antipolis (1965 - 2019) (UNS), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis (I3S), COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-COMUE Université Côte d'Azur (2015-2019) (COMUE UCA)-Centre National de la Recherche Scientifique (CNRS)-Université Côte d'Azur (UCA)-Centre National de la Recherche Scientifique (CNRS), Neuroimagerie: méthodes et applications (EMPENN), Institut National de la Santé et de la Recherche Médicale (INSERM)-Inria Rennes – Bretagne Atlantique, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-SIGNAL, IMAGE ET LANGAGE (IRISA-D6), Institut de Recherche en Informatique et Systèmes Aléatoires (IRISA), Université de Rennes (UR)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université de Bretagne Sud (UBS)-École normale supérieure - Rennes (ENS Rennes)-Institut National de Recherche en Informatique et en Automatique (Inria)-CentraleSupélec-Centre National de la Recherche Scientifique (CNRS)-IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT)-Université de Rennes (UR)-Institut National des Sciences Appliquées - Rennes (INSA Rennes), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT)-Institut de Recherche en Informatique et Systèmes Aléatoires (IRISA), Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université de Bretagne Sud (UBS)-École normale supérieure - Rennes (ENS Rennes)-CentraleSupélec-Centre National de la Recherche Scientifique (CNRS)-IMT Atlantique (IMT Atlantique), Institut Mines-Télécom [Paris] (IMT)-Institut Mines-Télécom [Paris] (IMT), GIPSA - Vision and Brain Signal Processing (GIPSA-VIBS), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé (CREATIS), Université Claude Bernard Lyon 1 (UCBL), Université de Lyon-Université de Lyon-Institut National des Sciences Appliquées de Lyon (INSA Lyon), and Université de Lyon-Institut National des Sciences Appliquées (INSA)-Institut National des Sciences Appliquées (INSA)-Université Jean Monnet - Saint-Étienne (UJM)-Institut National de la Santé et de la Recherche Médicale (INSERM)-Centre National de la Recherche Scientifique (CNRS)
Subjects: [SDV.IB.IMA]Life Sciences [q-bio]/Bioengineering/Imaging
Abstract: International audience; Au cours des dernières décennies, les progrès technologiques ont débouché sur le développement de nouveaux capteurs et des moyens de calculs très performants permettant l’acquisition et le traitement très rapide, voire en temps réel, de très grandes quantités de signaux et d’images. Ces avancées ont impacté de nombreux domaines, aboutissant dans le secteur de la santé à des techniques révolutionnaires pour visualiser le corps.Ces nouvelles imageries pour la santé incluent des méthodes de visualisation des organes jusqu’à l’échelle cellulaire ou moléculaire, des méthodes d’évaluation permettant de vérifier le bon fonctionnement des organes, et des méthodes de suivi interventionnel permettant de guider les gestes du praticien hospitalier ou du chirurgien.L’apport des sciences du numérique et de l’intelligence artificielle est essentiel pour ces nouvelles technologies, et permet au corps médical d’analyser rapidement de très grandes quantités d’images, de les traiter conjointement lorsqu’elles proviennent de différents dispositifs et d’exploiter les données des praticiens du monde entier. Mais l’utilisation de ces nouvelles technologies pose également des problèmes sociétaux, notamment de formation des soignants et d’acceptation, tant par le patient que par le praticien. Enfin, le partage de données médicales soulève des questions d’éthique et de préservation de la vie privée.Abondamment illustré, en 17 articles indépendants, Le corps en images aborde toutes ces questions et présente les progrès récents et les perspectives de l’imagerie médicale, un domaine qui nous concerne tous.
Published: 2022

42. Siamese Transformer Network for Hyperspectral Image Target Detection

Author: Weiqiang Rao, Lianru Gao, Ying Qu, Xu Sun, Bing Zhang, Jocelyn Chanussot, Chinese Academy of Sciences [Beijing] (CAS), University of Chinese Academy of Sciences [Beijing] (UCAS), The University of Tennessee [Knoxville], GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Hyperspectral imaging, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, Object detection, Transformers, General Earth and Planetary Sciences, Feature extraction, Training, Detectors, Electrical and Electronic Engineering, Signal processing algorithms
Abstract: International audience; Hyperspectral target detection can be described as locating targets of interest within a hyperspectral image based on prior information of targets. The complexity of actual scenes limits the performance of traditional statistical methods that rely on model assumptions, and traditional machine learning methods rely on mapping functions with limited complexity. To address these problems, we propose a Siamese transformer network for hyperspectral image target detection (STTD). The contribution of this article is threefold. First, we propose a novel method of constructing training samples using only the image itself and the limited prior information, which is suitable for target detection based on the Siamese network framework. Second, the Siamese network framework is utilized to solve the problem of similarity metric learning, i.e., make homogeneous features as close as possible and heterogeneous features as far as possible. Third, the most state-of-the-art network, transformer, is applied as the backbone of our proposed Siamese network to extract global features from spectra with long-range dependencies to achieve target detection. Furthermore, we make adaptive improvements to transformer for hyperspectral images. The proposed method shows its unique advantages in suppressing the background to a low level and highlighting the target with high probability. Experiments on five different datasets demonstrate the superiority of the proposed STTD as compared to the state-of-the-art.
Published: 2022

43. Analyse physiologique et acoustique des tekye dans le Tahrir iranien

Author: Castellengo, Michèle, During, Jean, Henrich Bernardoni, Nathalie, Institut Jean Le Rond d'Alembert (DALEMBERT), Sorbonne Université (SU)-Centre National de la Recherche Scientifique (CNRS), Centre de recherche en ethomusicologie (CREM), Laboratoire d'ethnologie et de sociologie comparative (LESC), Université Paris Nanterre (UPN)-Centre National de la Recherche Scientifique (CNRS)-Université Paris Nanterre (UPN)-Centre National de la Recherche Scientifique (CNRS), GIPSA - Analyse et Modification de l'homme en mouvement : biomécanique, cognition, vocologie (GIPSA-MOVE), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Société Française d'Acoustique, and Laboratoire de Mécanique et d'Acoustique
Subjects: [PHYS.MECA.ACOU]Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph]
Abstract: National audience; Une des caractéristiques les plus remarquables du chant classique persan est la technique du tahrir qui consiste à orner la ligne mélodique principale en produisant un ou plusieurs sauts brefs de la fréquence fondamentale vers l’aigu appelés tekye. Les caractéristiques acoustiques et physiologiques du tekye restent à mieux comprendre. Nous présentons d’une part, les résultats d’une étude de la physiologie laryngée par mesure électroglottographique conduite avec deux chanteurs iraniens professionnels et d’autre part, les relations entre l’analyse acoustique et les caractéristiques vibratoires laryngées des chanteurs pendant la réalisation musicale du tekye. L’analyse de la physiologie laryngée confirme le fait que le tekye est produit par un emprunt rapide (de 60 à 80 ms) au mécanisme vibratoire laryngé M2. [renvoi Bib] L'analyse spectrographique met en évidence la prédominance du deuxième harmonique contrôlé par les voyelles /o/ et /a/, qui assure l'homogénité vocale d'une succession de tekye.
Published: 2022

44. Séparation et reconstruction d’arêtes par analyse en chirplets d’ordres supérieurs

Author: Maison, Timothée, Silva, Fabrice, Henrich Bernardoni, Nathalie, Guillemain, Philippe, Laboratoire de Mécanique et d'Acoustique [Marseille] (LMA ), Aix Marseille Université (AMU)-École Centrale de Marseille (ECM)-Centre National de la Recherche Scientifique (CNRS), GIPSA - Analyse et Modification de l'homme en mouvement : biomécanique, cognition, vocologie (GIPSA-MOVE), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Société Française d'Acoustique, and Laboratoire de Mécanique et d'Acoustique
Subjects: [PHYS.MECA.ACOU]Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph]
Abstract: National audience; Dans le but de caractériser in vivo les résonances du conduit vocal et de quantifier leur rôle dans certaines productions vocales, un dispositif expérimental en développement s’appuie sur une excitation acoustique au niveau des lèvres par un sinus glissant linéaire (chirp) couplée à un microphone captant la pression aux lèvres ou à distance. Le signal mesuré est la somme du chirp filtré par le conduit vocal, d’un signal de voix (que l’on suppose harmonique et stationnaire) et de bruit éventuel. L’estimation des résonances du conduit vocal (en fréquence et facteur de qualité) demande alors une séparation des signaux permettant la reconstruction de la modulation d’amplitude du chirp. La connaissance a priori de la trajectoire du chirp dans le plan temps-fréquence et des hypothèses sur le signal voisé suggèrent l’utilisation de la transformée dite chirplet, généralisation de la transformée de Gabor intégrant la variation de fréquence instantanée (chirp rate), pour séparer les arêtes du plan temps- fréquence qui se croisent avec des pentes différentes. Dans un premier temps, l’amplitude du chirp est supposée localement constante sur la trame d’analyse avec des résultats intéressants en termes de séparation et de reconstruction du chirp. Puis, compte-tenu de la courte durée du chirp dans le contexte de la caractérisation du conduit vocal, nous cherchons à tirer parti de chirplets d’ordres supérieurs pour traiter les fortes modulations d’amplitude. Nous montrons que la famille des fonctions de Hermite (incluant la Gaussienne) constitue un choix pertinent de chirplets pour reconstruire fidèlement la modulation d’amplitude avec de bonnes propriétés de stabilité numérique. Finalement, l’utilisation de deux chirp rates complémentaires est proposé pour traiter le cas du croisement entre l’arête suivi (le chirp) et une harmonique de la voix. Le choix des valeurs des paramètres d’analyse est discuté, notamment celui de la taille de la fenêtre d’analyse.
Published: 2022

45. Précision et limites des mesures de résonance d’un conduit vocal par excitation acoustique aux lèvres : simulation numérique et étude sur maquettes

Author: Maison, Timothée, Allain, Baptiste, Silva, Fabrice, Guillemain, Philippe, Henrich Bernardoni, Nathalie, Laboratoire de Mécanique et d'Acoustique [Marseille] (LMA ), Aix Marseille Université (AMU)-École Centrale de Marseille (ECM)-Centre National de la Recherche Scientifique (CNRS), GIPSA - Analyse et Modification de l'homme en mouvement : biomécanique, cognition, vocologie (GIPSA-MOVE), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), Société Française d'Acoustique, and Laboratoire de Mécanique et d'Acoustique
Subjects: [PHYS.MECA.ACOU]Physics [physics]/Mechanics [physics]/Acoustics [physics.class-ph]
Abstract: National audience; La caractérisation des résonances du conduit vocal est indispensable pour comprendre l'acoustique de la voix. Les méthodes de mesure s’appuyant sur un signal d'excitation à large bande et une mesure de pression au niveau des lèvres offrent un bon compromis entre précision, rapidité et intrusivité. Nous abordons ici quelques conseils pratiques pour réaliser de telles mesures et obtenir des estimations précises des fréquences et des facteurs de qualité des résonances. La méthode de mesure est validée par comparaison avec des modèles numériques (lignes de transmission avec pertes visco-thermiques et rayonnement) et avec différentes mesures d’impédances, tout d’abord sur un simple tube métallique puis sur des maquettes simplifiées de voyelles. L'analyse modale (fréquences et facteurs de qualité) est réalisée avec la méthode des cercles de Kennelly exploitant à la fois l’amplitude et la phase des réponses en fréquence, robuste face au problème de proximité entre résonances et anti-résonances. D’autre part, étant donné les niveaux sonores émis par les chanteurs, nous expérimentons la possibilité d’éloigner le microphone (l’excitation restant au niveau des lèvres). On montre que le microphone peut être placé n'importe où dès lors que le rapport signal/bruit est suffisamment bon : sous hypothèse de rayonnement monopolaire, le modèle de mesure reste valable mais les estimations deviennent plus sensibles aux possibles interférences d'autres sources acoustiques. Les fréquences de résonance et les facteurs de qualité obtenus à partir de mesures avec un microphone jusqu'à 30 cm des lèvres s'écartent de moins de 0,2 % et 10 % respectivement des valeurs de référence (mesure aux lèvres), validant le principe d’une caractérisation avec un microphone éloigné des lèvres. Enfin, on discute les limites de la mesure avec excitation aux lèvres pour des configurations vocaliques avec constriction forte du conduit vocal.
Published: 2022

46. Keypoint-Based Local Descriptors for Target Recognition in SAR Images: A Comparative Analysis

Author: Ganggang Dong, Hongwei Liu, Jocelyn Chanussot, National University of Defense Technology [China], Duke University [Durham], GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: Synthetic aperture radar, comparative analysis, General Computer Science, Computer science, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, 0211 other engineering and technologies, pose configuration, 02 engineering and technology, law.invention, radar target recognition, raw intensity values, law, Radar imaging, holistic features, 0202 electrical engineering, electronic engineering, information engineering, Computer vision, Electrical and Electronic Engineering, Radar, Instrumentation, 021101 geological & geomatics engineering, business.industry, imaging parameters, SAR images, keypoint-based local descriptors, General Earth and Planetary Sciences, Clutter, 020201 artificial intelligence & image processing, Artificial intelligence, business, Articulation (phonetics), [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Though widely studied over the years, radar target recognition is still far from being solved. Most earlier works rely on holistic features or raw intensity values, which are sensitive to the real-world sources of variability such as changes of pose, configuration, and imaging parameters; articulation; and occlusion. To solve these problems, this article resorts to the local descriptors around keypoints.
Published: 2021

47. A Comparison of Mobile VR Display Running on an Ordinary Smartphone With Standard PC Display for P300-BCI Stimulus Presentation

Author: Grégoire Cattan, Cesar Mendoza, Marco Congedo, Anton Andreev, Interaction Homme Machine Technologie (IHMTEK), GIPSA - Vision and Brain Signal Processing (GIPSA-VIBS), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), and GIPSA-Services (GIPSA-Services)
Subjects: FOS: Computer and information sciences, Computer science, Computer Science - Human-Computer Interaction, Electroencephalography, [SCCO]Cognitive science, Computer Science - Graphics, 0302 clinical medicine, Human–computer interaction, P300, Virtual Reality VR, Réalité Virtuelle RV, Electroencéphalographie EEG, medicine.diagnostic_test, 05 social sciences, Brain-Computer Interfaces BCI, Interface Cerveau-Machine ICM, [INFO.INFO-GR]Computer Science [cs]/Graphics [cs.GR], Electroencephalograhy EEG, ACM: H.: Information Systems/H.5: INFORMATION INTERFACES AND PRESENTATION (e.g., HCI), Potentiel évoqué, [INFO.INFO-ES]Computer Science [cs]/Embedded Systems, [SCCO.COMP]Cognitive science/Computer science, Virtual reality, Stimulus (physiology), 050105 experimental psychology, Eeg synchronization, Human-Computer Interaction (cs.HC), Gaming, 03 medical and health sciences, InformationSystems_MODELSANDPRINCIPLES, Artificial Intelligence, medicine, Event-related potential ERP, 0501 psychology and cognitive sciences, Head-Mounted Devices HMD, Electrical and Electronic Engineering, Brain–computer interface, Casque de réalité virtuelle, Interface Cerveau-Ordinateur ICO, [SCCO.NEUR]Cognitive science/Neuroscience, Significant difference, Game, Graphics (cs.GR), Control and Systems Engineering, Mobile phone, Personal computer, Casque de Réalité Virtuelle Passif, Jeu vidéo, 030217 neurology & neurosurgery, Software, ACM: K.: Computing Milieux/K.8: PERSONAL COMPUTING/K.8.0: General/K.8.0.0: Games
Abstract: A brain-computer interface (BCI) based on electroencephalography (EEG) is a promising technology for enhancing virtual reality (VR) applications-in particular, for gaming. We focus on the so-called P300-BCI, a stable and accurate BCI paradigm relying on the recognition of a positive event-related potential (ERP) occurring in the EEG about 300 ms post-stimulation. We implemented a basic version of such a BCI displayed on an ordinary and affordable smartphone-based head-mounted VR device: that is, a mobile and passive VR system (with no electronic components beyond the smartphone). The mobile phone performed the stimuli presentation, EEG synchronization (tagging) and feedback display. We compared the ERPs and the accuracy of the BCI on the VR device with a traditional BCI running on a personal computer (PC). We also evaluated the impact of subjective factors on the accuracy. The study was within-subjects, with 21 participants and one session in each modality. No significant difference in BCI accuracy was found between the PC and VR systems, although the P200 ERP was significantly wider and larger in the VR system as compared to the PC system., Comment: IEEE Transactions on Games, Institute of Electrical and Electronics Engineers, In press
Published: 2021

48. Coupled Convolutional Neural Network With Adaptive Response Function Learning for Unsupervised Hyperspectral Super Resolution

Author: Danfeng Hong, Lianru Gao, Bing Zhang, Jocelyn Chanussot, Ke Zheng, Ximin Cui, Wenzhi Liao, Chinese Academy of Sciences [Beijing] (CAS), IMEC (IMEC), Catholic University of Leuven - Katholieke Universiteit Leuven (KU Leuven), Flemish Institute for Technological Research (VITO), German Aerospace Center (DLR), University of Chinese Academy of Sciences [Beijing] (UCAS), China University of Mining and Technology (CUMT), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), Université Grenoble Alpes (UGA), and ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019)
Subjects: FOS: Computer and information sciences, Point spread function, hyperspectral image, Computer science, Computer Vision and Pattern Recognition (cs.CV), Computer Science - Computer Vision and Pattern Recognition, Adaptive learning, 0211 other engineering and technologies, super-resolution, 02 engineering and technology, Convolutional neural network, Convolution, FOS: Electrical engineering, electronic engineering, information engineering, Electrical and Electronic Engineering, Image resolution, 021101 geological & geomatics engineering, autoencoder, coupled convolutional neural network, business.industry, Deep learning, Image and Video Processing (eess.IV), Hyperspectral imaging, Pattern recognition, Function (mathematics), Electrical Engineering and Systems Science - Image and Video Processing, Autoencoder, General Earth and Planetary Sciences, Artificial intelligence, business, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
Abstract: International audience; Due to the limitations of hyperspectral imaging systems, hyperspectral imagery (HSI) often suffers from poor spatial resolution, thus hampering many applications of the imagery. Hyperspectral super resolution refers to fusing HSI and MSI to generate an image with both high spatial and high spectral resolutions. Recently, several new methods have been proposed to solve this fusion problem, and most of these methods assume that the prior information of the point spread function (PSF) and spectral response function (SRF) are known. However, in practice, this information is often limited or unavailable. In this work, an unsupervised deep learning-based fusion method-HyCoNet-that can solve the problems in HSI-MSI fusion without the prior PSF and SRF information is proposed. HyCoNet consists of three coupled autoencoder nets in which the HSI and MSI are unmixed into endmembers and abundances based on the linear unmixing model. Two special convolutional layers are designed to act as a bridge that coordinates with the three autoencoder nets, and the PSF and SRF parameters are learned adaptively in the two convolution layers during the training process. Furthermore, driven by the joint loss function, the proposed method is straightforward and easily implemented in an end-to-end training manner. The experiments performed in the study demonstrate that the proposed method performs well and produces robust results for different data sets and arbitrary PSFs and SRFs.
Published: 2021

49. Reconstruction Error in Nonuniformly Sampled Approximately Sparse Signals

Author: Milos Dakovic, Cornel Ioana, Isidora Stankovic, Milos Brajovic, Ljubisa Stankovic, University of Montenegro (UCG), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), and Université Grenoble Alpes (UGA)
Subjects: Noise measurement, Computer science, 0211 other engineering and technologies, Nonuniform sampling, Jitter, 02 engineering and technology, Domain (software engineering), symbols.namesake, [INFO.INFO-TS]Computer Science [cs]/Signal and Image Processing, 0202 electrical engineering, electronic engineering, information engineering, Range (statistics), Random variables, Electrical and Electronic Engineering, 021101 geological & geomatics engineering, Sensors, Sampling (statistics), Indexes, 020206 networking & telecommunications, Geotechnical Engineering and Engineering Geology, Discrete Fourier transforms, Noise, Compressed sensing, Fourier transform, symbols, Algorithm, Random variable
Abstract: International audience; With its aim to reduce the amount of sensed data and to improve the energy efficiency, compressive sensing (CS) is recently witnessing a growing research interest in remote-sensing applications. The Fourier transform domain plays a significant role as a signal-processing tool and the sparsity domain for the CS-reconstruction methods. A generalized expression for the error in the reconstruction of nonuniformly sampled, approximately sparse, or nonsparse, noisy signals in the Fourier domain is presented in this letter. This expression holds for a wide range of practically important nonuniform signal-sampling strategies, covering the uniform and completely random sampling as the special cases. Additive noise and noise-folding effects are included in the analysis. Statistical examples and two real-world examples validate the presented theory.
Published: 2021

50. Multimodal remote sensing benchmark datasets for land cover classification with a shared and specific feature learning model

Author: Hong, Danfeng, Hu, Jingliang, Yao, Jing, Chanussot, Jocelyn, Zhu, Xiao Xiang, German Aerospace Center (DLR), Technical University of Munich (TUM), Chinese Academy of Sciences [Beijing] (CAS), Laboratoire Jean Kuntzmann (LJK), Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA), GIPSA - Signal Images Physique (GIPSA-SIGMAPHY), GIPSA Pôle Sciences des Données (GIPSA-PSD), Grenoble Images Parole Signal Automatique (GIPSA-lab), Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Centre National de la Recherche Scientifique (CNRS)-Université Grenoble Alpes (UGA)-Institut polytechnique de Grenoble - Grenoble Institute of Technology (Grenoble INP ), Université Grenoble Alpes (UGA)-Grenoble Images Parole Signal Automatique (GIPSA-lab), ANR-19-P3IA-0003,MIAI,MIAI @ Grenoble Alpes(2019), Technische Universität Munchen - Université Technique de Munich [Munich, Allemagne] (TUM), Apprentissage de modèles à partir de données massives (Thoth), Inria Grenoble - Rhône-Alpes, and Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)-Laboratoire Jean Kuntzmann (LJK)
Subjects: FOS: Computer and information sciences, Multispectral, Computer Vision and Pattern Recognition (cs.CV), Computer Science - Computer Vision and Pattern Recognition, Feature learning, Remote sensing, Classification, HySpex, Article, DSM, Hyperspectral, Multimodal, Benchmark datasets, Shared features, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Land cover mapping, SAR, Specific features
Abstract: International audience; As remote sensing (RS) data obtained from different sensors become available largely and openly, multimodal data processing and analysis techniques have been garnering increasing interest in the RS and geoscience community. However, due to the gap between different modalities in terms of imaging sensors, resolutions, and contents, embedding their complementary information into a consistent, compact, accurate, and discriminative representation, to a great extent, remains challenging. To this end, we propose a shared and specific feature learning (S2FL) model. S2FL is capable of decomposing multimodal RS data into modality-shared and modality-specific components, enabling the information blending of multi-modalities more effectively, particularly for heterogeneous data sources. Moreover, to better assess multimodal baselines and the newly-proposed S2FL model, three multimodal RS benchmark datasets, i.e., Houston2013 – hyperspectral and multispectral data, Berlin – hyperspectral and synthetic aperture radar (SAR) data, Augsburg – hyperspectral, SAR, and digital surface model (DSM) data, are released and used for land cover classification. Extensive experiments conducted on the three datasets demonstrate the superiority and advancement of our S2FL model in the task of land cover classification in comparison with previously-proposed state-of-the-art baselines. Furthermore, the baseline codes and datasets used in this paper will be made available freely at https://github.com/danfenghong/ISPRS_S2FL.
Published: 2021

Catalog

Books, media, physical & digital resources

See catalog results

Searchworks

Select search scope, currently: Articles Catalog books, media & more in Jio Institute collections Articles journal articles & other e-resources

Search

Search Constraints

Refine your results

Search Limiters

Topic

Publication Year Range

Language

Journal

Database

Publisher

370 results on '"GIPSA Pôle Sciences des Données (GIPSA-PSD)"'

Search Results

Catalog

Select search scope, currently: Articles

Catalog

books, media & more in Jio Institute collections

Articles

journal articles & other e-resources