Search

Your search keyword '"Bouthemy, Patrick"' showing total 312 results
Start Over Author "Bouthemy, Patrick"

Refine your results

more »

more »

more »

more »
312 results on '"Bouthemy, Patrick"'

1. Segmenting the motion components of a video: A long-term unsupervised model

Author

Meunier, Etienne and Bouthemy, Patrick

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Human beings have the ability to continuously analyze a video and immediately extract the motion components. We want to adopt this paradigm to provide a coherent and stable motion segmentation over the video sequence. In this perspective, we propose a novel long-term spatio-temporal model operating in a totally unsupervised way. It takes as input the volume of consecutive optical flow (OF) fields, and delivers a volume of segments of coherent motion over the video. More specifically, we have designed a transformer-based network, where we leverage a mathematically well-founded framework, the Evidence Lower Bound (ELBO), to derive the loss function. The loss function combines a flow reconstruction term involving spatio-temporal parametric motion models combining, in a novel way, polynomial (quadratic) motion models for the spatial dimensions and B-splines for the time dimension of the video sequence, and a regularization term enforcing temporal consistency on the segments. We report experiments on four VOS benchmarks, demonstrating competitive quantitative results, while performing motion segmentation on a whole sequence in one go. We also highlight through visual results the key contributions on temporal consistency brought by our method.

Published
2023

2. EM-driven unsupervised learning for efficient motion segmentation

Author

Meunier, Etienne, Badoual, Anaïs, and Bouthemy, Patrick

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

In this paper, we present a CNN-based fully unsupervised method for motion segmentation from optical flow. We assume that the input optical flow can be represented as a piecewise set of parametric motion models, typically, affine or quadratic motion models. The core idea of our work is to leverage the Expectation-Maximization (EM) framework in order to design in a well-founded manner a loss function and a training procedure of our motion segmentation neural network that does not require either ground-truth or manual annotation. However, in contrast to the classical iterative EM, once the network is trained, we can provide a segmentation for any unseen optical flow field in a single inference step and without estimating any motion models. We investigate different loss functions including robust ones and propose a novel efficient data augmentation technique on the optical flow field, applicable to any network taking optical flow as input. In addition, our method is able by design to segment multiple motions. Our motion segmentation network was tested on four benchmarks, DAVIS2016, SegTrackV2, FBMS59, and MoCA, and performed very well, while being fast at test time., Comment: Accepted to : IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

Published
2022
Full Text
View/download PDF

3. Learning how to be robust: Deep polynomial regression

Author

Perez-Rua, Juan-Manuel, Crivelli, Tomas, Bouthemy, Patrick, and Perez, Patrick

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Polynomial regression is a recurrent problem with a large number of applications. In computer vision it often appears in motion analysis. Whatever the application, standard methods for regression of polynomial models tend to deliver biased results when the input data is heavily contaminated by outliers. Moreover, the problem is even harder when outliers have strong structure. Departing from problem-tailored heuristics for robust estimation of parametric models, we explore deep convolutional neural networks. Our work aims to find a generic approach for training deep regression models without the explicit need of supervised annotation. We bypass the need for a tailored loss function on the regression parameters by attaching to our model a differentiable hard-wired decoder corresponding to the polynomial operation at hand. We demonstrate the value of our findings by comparing with standard robust regression methods. Furthermore, we demonstrate how to use such models for a real computer vision problem, i.e., video stabilization. The qualitative and quantitative experiments show that neural networks are able to learn robustness for general polynomial regression, with results that well overpass scores of traditional robust estimation methods., Comment: 18 pages, conference

Published
2018

4. Tubelets: Unsupervised action proposals from spatiotemporal super-voxels

Author

Jain, Mihir, van Gemert, Jan, Jégou, Hervé, Bouthemy, Patrick, and Snoek, Cees G. M.

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

This paper considers the problem of localizing actions in videos as a sequences of bounding boxes. The objective is to generate action proposals that are likely to include the action of interest, ideally achieving high recall with few proposals. Our contributions are threefold. First, inspired by selective search for object proposals, we introduce an approach to generate action proposals from spatiotemporal super-voxels in an unsupervised manner, we call them Tubelets. Second, along with the static features from individual frames our approach advantageously exploits motion. We introduce independent motion evidence as a feature to characterize how the action deviates from the background and explicitly incorporate such motion information in various stages of the proposal generation. Finally, we introduce spatiotemporal refinement of Tubelets, for more precise localization of actions, and pruning to keep the number of Tubelets limited. We demonstrate the suitability of our approach by extensive experiments for action proposal quality and action localization on three public datasets: UCF Sports, MSR-II and UCF101. For action proposal quality, our unsupervised proposals beat all other existing approaches on the three datasets. For action localization, we show top performance on both the trimmed videos of UCF Sports and UCF101 as well as the untrimmed videos of MSR-II., Comment: submitted to International Journal of Computer Vision

Published
2016

5. Aggregation of local parametric candidates with exemplar-based occlusion handling for optical flow

Author

Fortun, Denis, Bouthemy, Patrick, and Kervrann, Charles

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Handling all together large displacements, motion details and occlusions remains an open issue for reliable computation of optical flow in a video sequence. We propose a two-step aggregation paradigm to address this problem. The idea is to supply local motion candidates at every pixel in a first step, and then to combine them to determine the global optical flow field in a second step. We exploit local parametric estimations combined with patch correspondences and we experimentally demonstrate that they are sufficient to produce highly accurate motion candidates. The aggregation step is designed as the discrete optimization of a global regularized energy. The occlusion map is estimated jointly with the flow field throughout the two steps. We propose a generic exemplar-based approach for occlusion filling with motion vectors. We achieve state-of-the-art results in computer vision benchmarks, with particularly significant improvements in the case of large displacements and occlusions., Comment: Submission,IEEE Transactions on Image Processing (2014)

Published
2014

9. Sparse Aggregation Framework for Optical Flow Estimation

Author

Fortun, Denis, Bouthemy, Patrick, Kervrann, Charles, Hutchison, David, Series editor, Kanade, Takeo, Series editor, Kittler, Josef, Series editor, Kleinberg, Jon M., Series editor, Mattern, Friedemann, Series editor, Mitchell, John C., Series editor, Naor, Moni, Series editor, Pandu Rangan, C., Series editor, Steffen, Bernhard, Series editor, Terzopoulos, Demetri, Series editor, Tygar, Doug, Series editor, Weikum, Gerhard, Series editor, Aujol, Jean-François, editor, Nikolova, Mila, editor, and Papadakis, Nicolas, editor

Published
2015
Full Text
View/download PDF

16. Simultaneous Motion Detection and Background Reconstruction with a Mixed-State Conditional Markov Random Field

Author

Crivelli, Tomás, Piriou, Gwenaelle, Bouthemy, Patrick, Cernuschi-Frías, Bruno, Yao, Jian-feng, Hutchison, David, Series editor, Kanade, Takeo, Series editor, Kittler, Josef, Series editor, Kleinberg, Jon M., Series editor, Mattern, Friedemann, Series editor, Mitchell, John C., Series editor, Naor, Moni, Series editor, Nierstrasz, Oscar, Series editor, Pandu Rangan, C., Series editor, Steffen, Bernhard, Series editor, Sudan, Madhu, Series editor, Terzopoulos, Demetri, Series editor, Tygar, Doug, Series editor, Vardi, Moshe Y., Series editor, Weikum, Gerhard, Series editor, Forsyth, David, editor, Torr, Philip, editor, and Zisserman, Andrew, editor

Published
2008
Full Text
View/download PDF

17. Rank-Test Similarity Measure Between Video Segments for Local Descriptors

Author

Lehmann, Alain, Bouthemy, Patrick, Yao, Jian-Feng, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Rangan, C. Pandu, editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Marchand-Maillet, Stéphane, editor, Bruno, Eric, editor, Nürnberger, Andreas, editor, and Detyniecki, Marcin, editor

Published
2007
Full Text
View/download PDF

18. A General Principled Method for Image Similarity Validation

Author

Cao, Frédéric, Bouthemy, Patrick, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Rangan, C. Pandu, editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Doug, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Marchand-Maillet, Stéphane, editor, Bruno, Eric, editor, Nürnberger, Andreas, editor, and Detyniecki, Marcin, editor

Published
2007
Full Text
View/download PDF

19. Motion-Based Segmentation of Transparent Layers in Video Sequences

Author

Auvray, Vincent, Bouthemy, Patrick, Liénard, Jean, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Gunsel, Bilge, editor, Jain, Anil K., editor, Tekalp, A. Murat, editor, and Sankur, Bülent, editor

Published
2006
Full Text
View/download PDF

20. Kernel-Based Robust Tracking for Objects Undergoing Occlusion

Author

Babu, R. Venkatesh, Pérez, Patrick, Bouthemy, Patrick, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Narayanan, P. J., editor, Nayar, Shree K., editor, and Shum, Heung-Yeung, editor

Published
2006
Full Text
View/download PDF

21. Multiresolution Parametric Estimation of Transparent Motions and Denoising of Fluoroscopic Images

Author

Auvray, Vincent, Liénard, Jean, Bouthemy, Patrick, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Duncan, James S., editor, and Gerig, Guido, editor

Published
2005
Full Text
View/download PDF

22. Adaptive Spatio-Temporal Restoration for 4D Fluorescence Microscopic Imaging

Author

Boulanger, Jérôme, Kervrann, Charles, Bouthemy, Patrick, Hutchison, David, editor, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Duncan, James S., editor, and Gerig, Guido, editor

Published
2005
Full Text
View/download PDF

24. A Geometrical Key-Frame Selection Method Exploiting Dominant Motion Estimation in Video

Author

Fauvet, Brigitte, Bouthemy, Patrick, Gros, Patrick, Spindler, Fabien, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Naor, Moni, editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Enser, Peter, editor, Kompatsiaris, Yiannis, editor, O’Connor, Noel E., editor, Smeaton, Alan F., editor, and Smeulders, Arnold W. M., editor

Published
2004
Full Text
View/download PDF

25. Extraction of Semantic Dynamic Content from Videos with Probabilistic Motion Models

Author

Piriou, Gwenaëlle, Bouthemy, Patrick, Yao, Jian-Feng, Kanade, Takeo, editor, Kittler, Josef, editor, Kleinberg, Jon M., editor, Mattern, Friedemann, editor, Mitchell, John C., editor, Nierstrasz, Oscar, editor, Pandu Rangan, C., editor, Steffen, Bernhard, editor, Sudan, Madhu, editor, Terzopoulos, Demetri, editor, Tygar, Dough, editor, Vardi, Moshe Y., editor, Weikum, Gerhard, editor, Pajdla, Tomáš, editor, and Matas, Jiří, editor

Published
2004
Full Text
View/download PDF

28. Apprentissage profond non supervisé fondé sur l'algorithme EM pour la segmentation du mouvement

Author

Meunier, Etienne, Badoual, Anaïs, Bouthemy, Patrick, Meunier, Etienne, Space-timE RePresentation, Imaging and cellular dynamics of molecular COmplexes (SERPICO), Inria Rennes – Bretagne Atlantique, and Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria)

Subjects
optical flow, [INFO.INFO-CV] Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], flot optique, algorithme EM, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], Segmentation du mouvement, Motion segmentation, EM algorithm, CNN, réseau de neurones
Abstract

This paper presents a CNN-based fully unsupervised method for motion segmentation from optical flow. We assume that the input optical flow can be represented as a piecewise set of parametric motion models, typically, affine or quadratic motion models. The core idea of this work is to leverage the Expectation-Maximization (EM) framework. It enables us to design in a well-founded manner the loss function and the training procedure of our motion segmentation neural network. However, in contrast to the classical iterative EM, once the network is trained, we can provide a segmentation for any unseen optical flow field in a single inference step, with no dependence on the initialization of the motion model parameters since they are not estimated in the inference stage. Our method outperforms comparable unsupervised methods and is very efficient., Cet article présente une méthode entièrement non supervisée basée sur un réseau neuronal convolutionnel pour la segmentation du mouvement à partir du flot optique. Nous supposons que le flot optique d'entrée peut être représenté par un ensemble de modèles de mouvement paramétriques, typiquement des modèles de mouvement affines ou quadratiques. L'idée centrale de ce travail est d'exploiter le cadre de l'Espérance-Maximisation (EM). Il nous permet de concevoir d'une manière bien fondée la fonction de perte et la procédure d'apprentissage de notre réseau de neurones de segmentation du mouvement. Contrairement à la méthode EM itérative classique, une fois le réseau entraîné, nous pouvons fournir une segmentation pour tout nouveau champ de flot optique en une seule étape d'inférence et sans avoir avoir à estimer de modèles de mouvement. Notre méthode surpasse les méthodes non supervisées comparables et est très efficace en temps de calcul. Mots Clés Segmentation du mouvement, flot optique, réseau de neurones, algorithme EM.

Published
2022

36. Localisation de mouvements saillants dans des cartes de flot optique par l'interprétation d'un réseau de classification

Author

Meunier, Etienne, Bouthemy, Patrick, Space-timE RePresentation, Imaging and cellular dynamics of molecular COmplexes (SERPICO), Inria Rennes – Bretagne Atlantique, Institut National de Recherche en Informatique et en Automatique (Inria)-Institut National de Recherche en Informatique et en Automatique (Inria), and Centre National de la Recherche Scientifique [CNRS]

Subjects
cartes d'attention, Saillance de mouvement, Apprentissage non supervisé, [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV], [INFO]Computer Science [cs], CNN
Abstract

International audience; Cet article s’intéresse à la localisation des mouvements saillants dans les images successives d’une séquence vidéo. Un mouvement saillant est un mouvement se démarquant de son contexte environnant. Notre méthode s’appuie sur l’interprétation, pour chaque instant de la séquence, d’un réseau convolutionnel de classification dont l’entrée est constituée du flot optique. Cette classification porte sur la présence ou non de mouvements saillants dans l’image. En combinant la carte d’interprétation déduite du réseau et une segmentation du flot optique, nous pouvons détourer précisément les mouvements saillants dans l’image et estimer leur degré de saillance. Un atout important de notre méthode est qu’elle ne nécessite pas de cartes de segmentation annotées pour l’entraînement du réseau. Les résultats expérimentaux et la comparaison avec les méthodes existantes démontrent la performance de notre méthode sur une variété de vidéos

Published
2021

43. The derivation of qualitative information in motion analysis

Author

François, Edouard, Bouthemy, Patrick, Goos, G., editor, Hartmanis, J., editor, Barstow, D., editor, Brauer, W., editor, Brinch Hansen, P., editor, Gries, D., editor, Luckham, D., editor, Moler, C., editor, Pnueli, A., editor, Seegmüller, G., editor, Stoer, J., editor, Wirth, N., editor, and Faugeras, O., editor

Published
1990
Full Text
View/download PDF

44. EM-Driven Unsupervised Learning for Efficient Motion Segmentation

Author

Meunier, Etienne, Badoual, Anais, and Bouthemy, Patrick

Abstract

In this paper, we present a CNN-based fully unsupervised method for motion segmentation from optical flow. We assume that the input optical flow can be represented as a piecewise set of parametric motion models, typically, affine or quadratic motion models. The core idea of our work is to leverage the Expectation-Maximization (EM) framework in order to design in a well-founded manner a loss function and a training procedure of our motion segmentation neural network that does not require either ground-truth or manual annotation. However, in contrast to the classical iterative EM, once the network is trained, we can provide a segmentation for any unseen optical flow field in a single inference step and without estimating any motion models. We investigate different loss functions including robust ones and propose a novel efficient data augmentation technique on the optical flow field, applicable to any network taking optical flow as input. In addition, our method is able by design to segment multiple motions. Our motion segmentation network was tested on four benchmarks, DAVIS2016, SegTrackV2, FBMS59, and MoCA, and performed very well, while being fast at test time.

Published
2023
Full Text
View/download PDF

48. A statistical video content recognition method using invariant features on object trajectories

Author

Hervieu, Alexandre, Bouthemy, Patrick, and Le Cadre, Jean-Pierre

Subjects
Markov processes -- Evaluation, Motion perception (Vision) -- Technology application, Object recognition (Computers) -- Evaluation, Pattern recognition -- Evaluation, Signal processing -- Methods, Automatic classification -- Methods, Video equipment -- Research, Technology application, Digital signal processor, Business, Computers, Electronics, Electronics and electrical industries
Abstract

This work is dedicated to a statistical trajectory-based approach addressing two issues related to dynamic video content understanding: recognition of events and detection of unexpected events. Appropriate local differential features combining curvature and motion magnitude are defined and robustly computed on the motion trajectories in the image sequence. These features are invariant to image translation, in-the-plane rotation and spatial scaling. The temporal causality of the features is then captured by Hidden Markov Models dedicated to trajectory description, whose states are properly quantized values. The similarity between trajectories is expressed by exploiting this quantization-based HMM framework. Moreover statistical techniques have been developed for parameter estimations. Evaluations of the method have been conducted on several data sets including real trajectories obtained from sport videos, especially Formula One and ski TV program. The novel method compares favorably with other methods including feature histogram comparisons, HMM/GMM modeling and SVM classification. Index Terms--Hidden Markov models, motion analysis, pattern classification, video signal processing.

Published
2008

Catalog

Books, media, physical & digital resources
See catalog results