Search

Your search keyword '"Francini, Gianluca"' showing total 35 results
Start Over Author "Francini, Gianluca" Search Limiters Peer Reviewed
35 results on '"Francini, Gianluca"'

2. Feature Fusion for Robust Patch Matching With Compact Binary Descriptors

Author

Migliorati, Andrea, Fiandrotti, Attilio, Francini, Gianluca, Lepsoy, Skjalg, and Leonardi, Riccardo

Subjects
Computer Science - Computer Vision and Pattern Recognition, Computer Science - Machine Learning
Abstract

This work addresses the problem of learning compact yet discriminative patch descriptors within a deep learning framework. We observe that features extracted by convolutional layers in the pixel domain are largely complementary to features extracted in a transformed domain. We propose a convolutional network framework for learning binary patch descriptors where pixel domain features are fused with features extracted from the transformed domain. In our framework, while convolutional and transformed features are distinctly extracted, they are fused and provided to a single classifier which thus jointly operates on convolutional and transformed features. We experiment at matching patches from three different datasets, showing that our feature fusion approach outperforms multiple state-of-the-art approaches in terms of accuracy, rate, and complexity., Comment: MMSP 2018 - IEEE 20th International Workshop on Multimedia Signal Processing - August 29-31 2018, Vancouver, Canada

Published
2019

4. Learning Sparse Neural Networks via Sensitivity-Driven Regularization

Author

Tartaglione, Enzo, Lepsøy, Skjalg, Fiandrotti, Attilio, and Francini, Gianluca

Subjects
Computer Science - Machine Learning, Condensed Matter - Disordered Systems and Neural Networks, Statistics - Machine Learning
Abstract

The ever-increasing number of parameters in deep neural networks poses challenges for memory-limited applications. Regularize-and-prune methods aim at meeting these challenges by sparsifying the network weights. In this context we quantify the output sensitivity to the parameters (i.e. their relevance to the network output) and introduce a regularization term that gradually lowers the absolute value of parameters with low sensitivity. Thus, a very large fraction of the parameters approach zero and are eventually set to zero by simple thresholding. Our method surpasses most of the recent techniques both in terms of sparsity and error rates. In some cases, the method reaches twice the sparsity obtained by other techniques at equal error rates.

Published
2018

5. Fast Training of Convolutional Neural Networks via Kernel Rescaling

Author

de Gusmão, Pedro Porto Buarque, Francini, Gianluca, Lepsøy, Skjalg, and Magli, Enrico

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Training deep Convolutional Neural Networks (CNN) is a time consuming task that may take weeks to complete. In this article we propose a novel, theoretically founded method for reducing CNN training time without incurring any loss in accuracy. The basic idea is to begin training with a pre-train network using lower-resolution kernels and input images, and then refine the results at the full resolution by exploiting the spatial scaling property of convolutions. We apply our method to the ImageNet winner OverFeat and to the more recent ResNet architecture and show a reduction in training time of nearly 20% while test set accuracy is preserved in both cases.

Published
2016

22. Proposal for a definition of CDVS engine APIs

Author

Balestri, Massimo, Bober, Miroslaw, Chiariglione, Leonardo, Cordara, Giovanni, Francini, Gianluca, Gibellino, Diego, Lavric, Traian, Messina, Alberto, Moran Burgos, Francisco, Paschalakis, Stavros, Preda, Marius, Telecom Italia, University of Surrey (UNIS), CEDEO.net (CEDEO.net), Huawei Technologies [Dusseldorf], Département Advanced Research And Techniques For Multidimensional Imaging Systems (ARTEMIS), Institut Mines-Télécom [Paris] (IMT)-Télécom SudParis (TSP), Radiotelevisione Italiana (RAI), Universidad Politécnica de Madrid (UPM), Visual Atoms (.), Dépt. ARTEMIS (Institut Mines-Télécom-Télécom SudParis), Telecom Italia (.), University of Surrey (University of Surrey), Radiotelevisione Italiana (.), Huawei Technologies [Dusseldorf] (Huawei ), and Universidad Politécnica de Madrid (Universidad Politécnica de Madrid)

Subjects
CDVS, MPEG, [SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing, Visual object tracking
Abstract

108th MPEG meeting, Valencia, ES - April 2014 - ISO/IEC JTC 1/SC 29/WG 11 N14310

Published
2014

27. Deep Classifiers-Based License Plate Detection, Localization and Recognition on GPU-Powered Mobile Platform.

Author

Syed Tahir Hussain Rizvi, Patti, Denis, Björklund, Tomas, Cabodi, Gianpiero, and Francini, Gianluca

Subjects
GRAPHICS processing units, MOBILE communication systems, AUTOMOBILE license plates
Abstract

The realization of a deep neural architecture on a mobile platform is challenging, but can open up a number of possibilities for visual analysis applications. A neural network can be realized on a mobile platform by exploiting the computational power of the embedded GPU and simplifying the flow of a neural architecture trained on the desktop workstation or a GPU server. This paper presents an embedded platform-based Italian license plate detection and recognition system using deep neural classifiers. In this work, trained parameters of a highly precise automatic license plate recognition (ALPR) system are imported and used to replicate the same neural classifiers on a Nvidia Shield K1 tablet. A CUDA-based framework is used to realize these neural networks. The flow of the trained architecture is simplified to perform the license plate recognition in real-time. Results show that the tasks of plate and character detection and localization can be performed in real-time on a mobile platform by simplifying the flow of the trained architecture. However, the accuracy of the simplified architecture would be decreased accordingly. [ABSTRACT FROM AUTHOR]

Published
2017
Full Text
View/download PDF

28. Extended template for the Advanced Surveillance AF

Author

Andrade, Maria Teresa, Barbosa, Vítor, Carreras, Anna, Carvalho, Pedro, Cordara, Giovanni, Delgado, Jaime, Dogan, Safak, Dufaux, Frederic, Ebrahimi, Touradj, Francini, Gianluca, Gallego, Isabel, Goldmann, Lutz, Ivanov, Ivan, Minh, Thien Ha, Jin, Shan, Kodikara Arachchi, Hemantha, Mohammadi, Gelareh, Mrak, Marta, Pietrowcew, Adam, Rama, Toni, Rodríguez, Eva, Sikora, Thomas, and Tous, Rubén

Subjects
Advanced Surveillance, AF, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Data_CODINGANDINFORMATIONTHEORY
Abstract

MPEG meeting proposal.

Published
2009

29. Contribution to the Advanced Surveillance AF

Author

Andrade, Maria Teresa, Barbosa, Vítor, Carreras, Anna, Carvalho, Pedro, Cordara, Giovanni, Delgado, Jaime, Dogan, Safak, Dufaux, Frederic, Ebrahimi, Touradj, Francini, Gianluca, Gallego, Isabel, Goldmann, Lutz, Ivanov, Ivan, Minh, Thien Ha, Jin, Shan, Kodikara Arachchi, Hemantha, Mohammadi, Gelareh, Mrak, Marta, Pietrowcew, Adam, Rama, Toni, Rodríguez, Eva, Sikora, Thomas, and Tous, Rubén

Subjects
Advanced Surveillance, AF, ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION, Data_CODINGANDINFORMATIONTHEORY
Abstract

MPEG meeting proposal.

Published
2009

31. DynamicTV

Author

Antonelli, Fabrizio, primary, Francini, Gianluca, additional, Geymonat, Marina, additional, and Lepsøy, Skjalg, additional

Published
2009
Full Text
View/download PDF

34. A search and retrieval framework for the management of copyrighted audiovisual content

Author

Carreras, Anna, Cordara, Giovanni, Delgado, Jaime, Frederic Dufaux, Francini, Gianluca, Ha, Thien M., Rodriguez, Eva, and Tous, Ruben

Subjects
ComputingMilieux_LEGALASPECTSOFCOMPUTING
Abstract

This paper presents a search and retrieval framework that enables the management of Intellectual Property in the World Wide Web. This twofold framework helps users to detect digital rights infringements of their copyrighted content. In order to detect possible copyright infringments, first the system crawls the Web to search replicas of user’s images, and later evaluates if the copies have been taken respecting the terms stated by the owner. On the other hand, this framework also helps users in finding something interesting in the Web. It will provide copyrighted content to users according to their preferences and to intellectual property rights integrating search and retrieval with digital rights management tools.

Catalog

Books, media, physical & digital resources
See catalog results