1 code implementation • 12 Oct 2024 • Mustafa Shukor, Matthieu Cord
(2) Skipping computations during training can recover 97% of the original performance, even when skipping half of the blocks or removing 70% of the weights.
no code implementations • 18 Sep 2024 • Amaia Cardiel, Eloi Zablocki, Elias Ramzi, Oriane Siméoni, Matthieu Cord
In this work, we propose LLM-wrapper, a method for 'black-box' adaptation of VLMs for the REC task using Large Language Models (LLMs).
1 code implementation • 17 Sep 2024 • Yihong Xu, Victor Letzelter, Mickaël Chen, Éloi Zablocki, Matthieu Cord
Additionally, to compensate for limited performance, some approaches rely on training with a large set of hypotheses, requiring a post-selection step during inference to significantly reduce the number of predictions.
1 code implementation • 12 Sep 2024 • Yuan Yin, Pegah Khayatan, Éloi Zablocki, Alexandre Boulch, Matthieu Cord
Machine learning based autonomous driving systems often face challenges with safety-critical scenarios that are rare in real-world data, hindering their large-scale deployment.
no code implementations • 12 Jun 2024 • Jayneel Parekh, Pegah Khayatan, Mustafa Shukor, Alasdair Newson, Matthieu Cord
The elements of the learned dictionary correspond to our proposed concepts.
1 code implementation • 12 Jun 2024 • Yihong Xu, Éloi Zablocki, Alexandre Boulch, Gilles Puy, Mickael Chen, Florent Bartoccioni, Nermin Samet, Oriane Siméoni, Spyros Gidaris, Tuan-Hung Vu, Andrei Bursuc, Eduardo Valle, Renaud Marlet, Matthieu Cord
In end-to-end forecasting, the model must jointly detect and track from sensor data (cameras or LiDARs) the past trajectories of the different elements of the scene and predict their future locations.
1 code implementation • 5 Jun 2024 • Paul Couairon, Mustafa Shukor, Jean-Emmanuel Haugeard, Matthieu Cord, Nicolas Thome
Foundation models have emerged as powerful tools across various domains including language, vision, and multimodal tasks.
1 code implementation • 26 May 2024 • Mustafa Shukor, Matthieu Cord
Perceptual tokens (1) are easily distinguishable from textual ones inside LLMs, with significantly different representations, and complete translation to textual tokens does not exist.
no code implementations • 3 May 2024 • Hugo Laurençon, Léo Tronchon, Matthieu Cord, Victor Sanh
The growing interest in vision-language models (VLMs) has been driven by improvements in large language models and vision transformers.
Ranked #7 on Long-Context Understanding on MMNeedle
1 code implementation • 24 Apr 2024 • Folco Bertini Baldassini, Mustafa Shukor, Matthieu Cord, Laure Soulier, Benjamin Piwowarski
Large Language Models have demonstrated remarkable performance across various tasks, exhibiting the capacity to swiftly acquire new skills, such as through In-Context Learning (ICL) with minimal demonstration examples.
no code implementations • 8 Apr 2024 • Hugo Caselles-Dupré, Charles Mellerio, Paul Hérent, Alizée Lopez-Persem, Benoit Béranger, Mathieu Soularue, Pierre Fautrel, Gauthier Vernier, Matthieu Cord
The reconstruction of images observed by subjects from fMRI data collected during visual stimuli has made strong progress in the past decade, thanks to the availability of extensive fMRI datasets and advancements in generative models for image generation.
no code implementations • 29 Mar 2024 • Barbara Toniella Corradini, Mustafa Shukor, Paul Couairon, Guillaume Couairon, Franco Scarselli, Matthieu Cord
The pipeline is as follows: the image is passed to both a captioner model (i. e. BLIP) and a diffusion model (i. e., Stable Diffusion Model) to generate a text description and visual representation, respectively.
1 code implementation • 22 Mar 2024 • Lan Feng, Mohammadhossein Bahari, Kaouther Messaoud Ben Amor, Éloi Zablocki, Matthieu Cord, Alexandre Alahi
Vehicle trajectory prediction has increasingly relied on data-driven solutions, but their ability to scale to different data domains and the impact of larger dataset sizes on their generalization remain under-explored.
Ranked #1 on Trajectory Prediction on nuScenes (using extra training data)
no code implementations • 20 Mar 2024 • Théophane Vallaeys, Mustafa Shukor, Matthieu Cord, Jakob Verbeek
The abilities of large language models (LLMs) have recently progressed to unprecedented levels, paving the way to novel applications in a wide variety of areas.
1 code implementation • CVPR 2024 • Loick Chambon, Eloi Zablocki, Mickaël Chen, Florent Bartoccioni, Patrick Pérez, Matthieu Cord
To address this we propose PointBeV a novel sparse BeV segmentation model operating on sparse BeV cells instead of dense grids.
no code implementations • 21 Dec 2023 • Kaouther Messaoud, Kathrin Grosse, Mickael Chen, Matthieu Cord, Patrick Pérez, Alexandre Alahi
In this paper, we focus on backdoors - a security threat acknowledged in other fields but so far overlooked for trajectory prediction.
1 code implementation • 14 Dec 2023 • Thibaut Loiseau, Tuan-Hung Vu, Mickael Chen, Patrick Pérez, Matthieu Cord
Assessing the robustness of perception models to covariate shifts and their ability to detect out-of-distribution (OOD) inputs is crucial for safety-critical applications such as autonomous vehicles.
no code implementations • 11 Dec 2023 • Cédric Rommel, Victor Letzelter, Nermin Samet, Renaud Marlet, Matthieu Cord, Patrick Pérez, Eduardo Valle
Monocular 3D human pose estimation (3D-HPE) is an inherently ambiguous task, as a 2D pose in an image might originate from different possible 3D poses.
1 code implementation • 1 Dec 2023 • Loick Chambon, Eloi Zablocki, Mickael Chen, Florent Bartoccioni, Patrick Perez, Matthieu Cord
To address this, we propose PointBeV, a novel sparse BeV segmentation model operating on sparse BeV cells instead of dense grids.
no code implementations • 24 Nov 2023 • Eslam Abdelrahman, Liangbing Zhao, Vincent Tao Hu, Matthieu Cord, Patrick Perez, Mohamed Elhoseiny
Diffusion models break down the challenging task of generating data from high-dimensional distributions into a series of easier denoising steps.
1 code implementation • 1 Oct 2023 • Mustafa Shukor, Alexandre Rame, Corentin Dancette, Matthieu Cord
Based on our ICL study, (3) we push ICL further and propose new multimodal ICL variants such as; Multitask-ICL, Chain-of-Hindsight-ICL, and Self-Correcting-ICL.
no code implementations • 18 Sep 2023 • Asya Grechka, Guillaume Couairon, Matthieu Cord
For the specific task of image inpainting, the current guiding mechanism relies on copying-and-pasting the known regions from the input image at each denoising step.
no code implementations • 4 Sep 2023 • Cédric Rommel, Eduardo Valle, Mickaël Chen, Souhaiel Khalfaoui, Renaud Marlet, Matthieu Cord, Patrick Pérez
We present an innovative approach to 3D Human Pose Estimation (3D-HPE) by integrating cutting-edge diffusion models, which have revolutionized diverse fields, but are relatively unexplored in 3D-HPE.
1 code implementation • 30 Jul 2023 • Mustafa Shukor, Corentin Dancette, Alexandre Rame, Matthieu Cord
Our model is efficiently pretrained on many tasks, based on task balancing and multimodal curriculum learning.
1 code implementation • 18 Jul 2023 • Spyros Gidaris, Andrei Bursuc, Oriane Simeoni, Antonin Vobecky, Nikos Komodakis, Matthieu Cord, Patrick Pérez
Self-supervised learning can be used for mitigating the greedy needs of Vision Transformer networks for very large fully-annotated datasets.
no code implementations • ICCV 2023 • Guillaume Couairon, Marlène Careil, Matthieu Cord, Stéphane Lathuilière, Jakob Verbeek
Large-scale text-to-image diffusion models have significantly improved the state of the art in generative image modelling and allow for an intuitive and powerful user interface to drive the image generation process.
1 code implementation • NeurIPS 2023 • Hugo Laurençon, Lucile Saulnier, Léo Tronchon, Stas Bekman, Amanpreet Singh, Anton Lozhkov, Thomas Wang, Siddharth Karamcheti, Alexander M. Rush, Douwe Kiela, Matthieu Cord, Victor Sanh
Large multimodal models trained on natural documents, which interleave images and text, outperform models trained on image-text pairs on various multimodal benchmarks.
Ranked #14 on MMR total on MRR-Benchmark (using extra training data)
1 code implementation • 15 Jun 2023 • Yihong Xu, Loïck Chambon, Éloi Zablocki, Mickaël Chen, Alexandre Alahi, Matthieu Cord, Patrick Pérez
In fact, conventional forecasting methods are usually not trained nor tested in real-world pipelines (e. g., with upstream detection, tracking, and mapping modules).
1 code implementation • CVPR 2023 • Corentin Dancette, Spencer Whitehead, Rishabh Maheshwary, Ramakrishna Vedantam, Stefan Scherer, Xinlei Chen, Matthieu Cord, Marcus Rohrbach
In this work, we explore Selective VQA in both in-distribution (ID) and OOD scenarios, where models are presented with mixtures of ID and OOD data.
1 code implementation • NeurIPS 2023 • Alexandre Ramé, Guillaume Couairon, Mustafa Shukor, Corentin Dancette, Jean-Baptiste Gaya, Laure Soulier, Matthieu Cord
Foundation models are first pre-trained on vast unsupervised datasets and then fine-tuned on labeled data.
1 code implementation • ICCV 2023 • Mustafa Shukor, Corentin Dancette, Matthieu Cord
In this work, we propose to rather direct effort to efficient adaptations of existing models, and propose to augment Language Models with perception.
no code implementations • 24 Jan 2023 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
In this paper, we identity the uniformity of the quantization operator as a limitation of existing approaches, and propose a data-free non-uniform method.
1 code implementation • CVPR 2023 • Hugo Touvron, Matthieu Cord, Maxime Oquab, Piotr Bojanowski, Jakob Verbeek, Hervé Jégou
Given a neural network to be trained, for each sample we implicitly instantiate two altered networks, "submodels", with stochastic depth: i. e. activating only a subset of the layers and skipping others.
1 code implementation • 20 Dec 2022 • Alexandre Ramé, Kartik Ahuja, Jianyu Zhang, Matthieu Cord, Léon Bottou, David Lopez-Paz
In this paper, we thus propose model ratatouille, a new strategy to recycle the multiple fine-tunings of the same foundation model on diverse auxiliary tasks.
Ranked #19 on Domain Generalization on TerraIncognita
1 code implementation • 9 Dec 2022 • Hugo Touvron, Matthieu Cord, Maxime Oquab, Piotr Bojanowski, Jakob Verbeek, Hervé Jégou
We introduce submodel co-training, a regularization method related to co-training, self-distillation and stochastic depth.
Ranked #67 on Image Classification on ImageNet
1 code implementation • 8 Dec 2022 • Mustafa Shukor, Nicolas Thome, Matthieu Cord
Finally, we validate the generalization of the approach to other tasks (i. e, Food Recognition) and domains with structured text such as the Medical domain on the ROCO dataset.
Ranked #1 on Cross-Modal Retrieval on Recipe1M+
no code implementations • CVPR 2023 • Fabio Cermelli, Matthieu Cord, Arthur Douillard
%a In this paper, we present the first continual learning model capable of operating on both semantic and panoptic segmentation.
Ranked #2 on Continual Semantic Segmentation on ADE20K
1 code implementation • CVPR 2023 • Mehdi Zemni, Mickaël Chen, Éloi Zablocki, Hédi Ben-Younes, Patrick Pérez, Matthieu Cord
We conduct a set of experiments on counterfactual explanation benchmarks for driving scenes, and we show that our method can be adapted beyond classification, e. g., to explain semantic segmentation models.
4 code implementations • 20 Oct 2022 • Guillaume Couairon, Jakob Verbeek, Holger Schwenk, Matthieu Cord
Semantic image editing is an extension of image generation, with the additional constraint that the generated image should be as similar as possible to a given input image.
1 code implementation • 29 Aug 2022 • Mustafa Shukor, Guillaume Couairon, Matthieu Cord
Vision and Language Pretraining has become the prevalent approach for tackling multimodal downstream tasks.
no code implementations • 8 Jul 2022 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
The leap in performance in state-of-the-art computer vision methods is attributed to the development of deep neural networks.
1 code implementation • 27 Jun 2022 • Florent Bartoccioni, Éloi Zablocki, Andrei Bursuc, Patrick Pérez, Matthieu Cord, Karteek Alahari
Recent works in autonomous driving have widely adopted the bird's-eye-view (BEV) semantic map as an intermediate representation of the world.
Ranked #6 on Bird's-Eye View Semantic Segmentation on nuScenes
no code implementations • 22 May 2022 • Corentin Dancette, Matthieu Cord
Transformers have been matching deep convolutional networks for vision architectures in recent works.
no code implementations • 20 May 2022 • Rémy Sun, Alexandre Ramé, Clément Masson, Nicolas Thome, Matthieu Cord
To solve this issue, we propose a novel unmixing step in MIMO architectures that allows subnetworks to properly share features.
no code implementations • 20 May 2022 • Rémy Sun, Clément Masson, Gilles Hénaff, Nicolas Thome, Matthieu Cord
Deep architecture have proven capable of solving many tasks provided a sufficient amount of labeled data.
2 code implementations • 19 May 2022 • Alexandre Ramé, Matthieu Kirchmeyer, Thibaud Rahier, Alain Rakotomamonjy, Patrick Gallinari, Matthieu Cord
Standard neural networks struggle to generalize under distribution shifts in computer vision.
1 code implementation • 25 Apr 2022 • Antoine Saporta, Arthur Douillard, Tuan-Hung Vu, Patrick Pérez, Matthieu Cord
Unsupervised Domain Adaptation (UDA) is a transfer learning task which aims at training on an unlabeled target domain by leveraging a labeled source domain.
1 code implementation • 20 Apr 2022 • Mustafa Shukor, Guillaume Couairon, Asya Grechka, Matthieu Cord
We propose a new retrieval framework, T-Food (Transformer Decoders with MultiModal Regularization for Cross-Modal Food Retrieval) that exploits the interaction between modalities in a novel regularization scheme, while using only unimodal encoders at test time for efficient retrieval.
Ranked #3 on Cross-Modal Retrieval on Recipe1M
11 code implementations • 14 Apr 2022 • Hugo Touvron, Matthieu Cord, Hervé Jégou
Our evaluations on Image classification (ImageNet-1k with and without pre-training on ImageNet-21k), transfer learning and semantic segmentation show that our procedure outperforms by a large margin previous fully supervised training recipes for ViT.
Ranked #1 on Image Classification on ImageNet ReaL (Number of params metric)
no code implementations • 28 Mar 2022 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
Computationally expensive neural networks are ubiquitous in computer vision and solutions for efficient inference have drawn a growing attention in the machine learning community.
7 code implementations • 18 Mar 2022 • Hugo Touvron, Matthieu Cord, Alaaeldin El-Nouby, Jakob Verbeek, Hervé Jégou
(2) Fine-tuning the weights of the attention layers is sufficient to adapt vision transformers to a higher resolution and to other classification tasks.
Ranked #9 on Image Classification on CIFAR-10 (using extra training data)
1 code implementation • CVPR 2022 • Guillaume Couairon, Asya Grechka, Jakob Verbeek, Holger Schwenk, Matthieu Cord
Via the latent space of an auto-encoder, we iteratively transform the input image toward the target point, ensuring coherence and quality with a variety of novel regularization terms.
5 code implementations • 27 Dec 2021 • Hugo Touvron, Matthieu Cord, Alaaeldin El-Nouby, Piotr Bojanowski, Armand Joulin, Gabriel Synnaeve, Hervé Jégou
We show how to augment any convolutional network with an attention-based global map to achieve non-local reasoning.
Ranked #39 on Semantic Segmentation on ADE20K val
no code implementations • 6 Dec 2021 • Himalaya Jain, Tuan-Hung Vu, Patrick Pérez, Matthieu Cord
With the rapid advances in generative adversarial networks (GANs), the visual quality of synthesised scenes keeps improving, including for complex urban scenes with applications to automated driving.
no code implementations • 6 Dec 2021 • Guillaume Couairon, Matthieu Cord, Matthijs Douze, Holger Schwenk
We introduce the SIMAT dataset to evaluate the task of Image Retrieval with Multimodal queries.
no code implementations • NeurIPS 2021 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
Deep Neural Networks (DNNs) are ubiquitous in today's computer vision landscape, despite involving considerable computational costs.
1 code implementation • CVPR 2022 • Arthur Douillard, Alexandre Ramé, Guillaume Couairon, Matthieu Cord
Our strategy scales to a large number of tasks while having negligible memory and time overheads due to strict control of the parameters expansion.
Ranked #2 on Incremental Learning on ImageNet - 10 steps
1 code implementation • 17 Nov 2021 • Paul Jacob, Éloi Zablocki, Hédi Ben-Younes, Mickaël Chen, Patrick Pérez, Matthieu Cord
In this work, we address the problem of producing counterfactual explanations for high-quality images and complex scenes.
1 code implementation • NeurIPS 2021 • Thomas Fel, Remi Cadene, Mathieu Chalvidal, Matthieu Cord, David Vigouroux, Thomas Serre
We describe a novel attribution method which is grounded in Sensitivity Analysis and uses Sobol indices.
no code implementations • 30 Sep 2021 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
Pruning Deep Neural Networks (DNNs) is a prominent field of study in the goal of inference runtime acceleration.
no code implementations • 29 Sep 2021 • Charles Corbière, Marc Lafon, Nicolas Thome, Matthieu Cord, Patrick Perez
A crucial property of KLoS is to be a class-wise divergence measure built from in-distribution samples and to not require OOD training data, in contrast to current second-order uncertainty measures.
1 code implementation • 16 Sep 2021 • Hédi Ben-Younes, Éloi Zablocki, Mickaël Chen, Patrick Pérez, Matthieu Cord
Learning-based trajectory prediction models have encountered great success, with the promise of leveraging contextual information in addition to motion history.
1 code implementation • 8 Sep 2021 • Florent Bartoccioni, Éloi Zablocki, Patrick Pérez, Matthieu Cord, Karteek Alahari
In such a monocular setup, dense depth is obtained with either additional input from one or several expensive LiDARs, e. g., with 64 beams, or camera-only methods, which suffer from scale-ambiguity and infinite-depth problems.
2 code implementations • 7 Sep 2021 • Alexandre Rame, Corentin Dancette, Matthieu Cord
In this paper, we introduce a new regularization - named Fishr - that enforces domain invariance in the space of the gradients of the loss: specifically, the domain-level variances of gradients are matched across training domains.
Ranked #30 on Domain Generalization on TerraIncognita
1 code implementation • ICCV 2021 • Antoine Saporta, Tuan-Hung Vu, Matthieu Cord, Patrick Pérez
In this work, we address the task of unsupervised domain adaptation (UDA) for semantic segmentation in presence of multiple target domains: The objective is to train a single model that can handle all these domains at test time.
2 code implementations • 29 Jun 2021 • Arthur Douillard, Yifu Chen, Arnaud Dapogny, Matthieu Cord
classes predicted by the old model to deal with background shift and avoid catastrophic forgetting of the old classes.
Ranked #7 on Overlapped 15-1 on PASCAL VOC 2012
Class Incremental Learning Continual Semantic Segmentation +5
1 code implementation • CVPR 2021 • Guillaume Le Moing, Tuan-Hung Vu, Himalaya Jain, Patrick Pérez, Matthieu Cord
Despite the recent progress of generative adversarial networks (GANs) at synthesizing photo-realistic images, producing complex urban scenes remains a challenging problem.
no code implementations • 31 May 2021 • Edouard Yvinec, Arnaud Dapogny, Matthieu Cord, Kevin Bailly
Deep Neural Networks (DNNs) are ubiquitous in today's computer vision land-scape, despite involving considerable computational costs.
15 code implementations • NeurIPS 2021 • Hugo Touvron, Piotr Bojanowski, Mathilde Caron, Matthieu Cord, Alaaeldin El-Nouby, Edouard Grave, Gautier Izacard, Armand Joulin, Gabriel Synnaeve, Jakob Verbeek, Hervé Jégou
We present ResMLP, an architecture built entirely upon multi-layer perceptrons for image classification.
Ranked #1 on Image Classification on Certificate Verification
1 code implementation • ICCV 2021 • Corentin Dancette, Remi Cadene, Damien Teney, Matthieu Cord
We use this new evaluation in a large-scale study of existing approaches for VQA.
Ranked #1 on Visual Question Answering (VQA) on VQA-CE
19 code implementations • ICCV 2021 • Hugo Touvron, Matthieu Cord, Alexandre Sablayrolles, Gabriel Synnaeve, Hervé Jégou
In particular, we investigate the interplay of architecture and optimization of such dedicated transformers.
Ranked #5 on Image Classification on Stanford Cars
1 code implementation • ICCV 2021 • Alexandre Rame, Remy Sun, Matthieu Cord
Recent strategies achieved ensembling "for free" by fitting concurrently diverse subnetworks inside a single base network.
Ranked #15 on Image Classification on Tiny ImageNet Classification
no code implementations • ICLR 2021 • Alexandre Rame, Matthieu Cord
Deep ensembles perform better than a single network thanks to the diversity among their members.
no code implementations • 13 Jan 2021 • Éloi Zablocki, Hédi Ben-Younes, Patrick Pérez, Matthieu Cord
The concept of explainability has several facets and the need for explainability is strong in driving, a safety-critical application.
34 code implementations • 23 Dec 2020 • Hugo Touvron, Matthieu Cord, Matthijs Douze, Francisco Massa, Alexandre Sablayrolles, Hervé Jégou
In this work, we produce a competitive convolution-free transformer by training on Imagenet only.
Ranked #4 on Efficient ViTs on ImageNet-1K (with DeiT-S)
3 code implementations • CVPR 2021 • Spyros Gidaris, Andrei Bursuc, Gilles Puy, Nikos Komodakis, Matthieu Cord, Patrick Pérez
With this in mind, we propose a teacher-student scheme to learn representations by training a convolutional net to reconstruct a bag-of-visual-words (BoW) representation of an image, given as input a perturbed version of that same image.
Ranked #18 on Semi-Supervised Image Classification on ImageNet - 1% labeled data (Top 5 Accuracy metric)
no code implementations • 11 Dec 2020 • Charles Corbière, Nicolas Thome, Antoine Saporta, Tuan-Hung Vu, Matthieu Cord, Patrick Pérez
In this paper, we introduce a novel target criterion for model confidence, namely the true class probability (TCP).
1 code implementation • 9 Dec 2020 • Hédi Ben-Younes, Éloi Zablocki, Patrick Pérez, Matthieu Cord
In this era of active development of autonomous vehicles, it becomes crucial to provide driving systems with the capacity to explain their decisions.
1 code implementation • 4 Dec 2020 • Taylor Mordan, Matthieu Cord, Patrick Pérez, Alexandre Alahi
By increasing the number of attributes jointly learned, we highlight an issue related to the scales of gradients, which arises in MTL with numerous tasks.
no code implementations • ICCV 2021 • Hugo Touvron, Alexandre Sablayrolles, Matthijs Douze, Matthieu Cord, Hervé Jégou
By jointly leveraging the coarse labels and the underlying fine-grained latent space, it significantly improves the accuracy of category-level retrieval methods.
Ranked #2 on Learning with coarse labels on cifar100
Fine-Grained Image Classification Learning with coarse labels +3
2 code implementations • CVPR 2021 • Arthur Douillard, Yifu Chen, Arnaud Dapogny, Matthieu Cord
classes predicted by the old model to deal with background shift and avoid catastrophic forgetting of the old classes.
Ranked #1 on Domain 11-5 on Cityscapes val
Class Incremental Learning Continual Semantic Segmentation +16
no code implementations • 13 Aug 2020 • Hugo Touvron, Matthijs Douze, Matthieu Cord, Hervé Jégou
We propose a simple architecture to address unpaired image-to-image translation tasks: style or class transfer, denoising, deblurring, deblocking, etc.
Ranked #1 on Image-to-Image Translation on vangogh2photo (Frechet Inception Distance metric)
1 code implementation • 24 Jun 2020 • Arthur Douillard, Eduardo Valle, Charles Ollion, Thomas Robert, Matthieu Cord
Continual learning aims to learn tasks sequentially, with (often severe) constraints on the storage of old learning samples, without suffering from catastrophic forgetting.
1 code implementation • 17 Jun 2020 • Corentin Dancette, Remi Cadene, Xinlei Chen, Matthieu Cord
First, we propose the Modifying Count Distribution (MCD) protocol, which penalizes models that over-rely on statistical shortcuts.
1 code implementation • 15 Jun 2020 • Antoine Saporta, Tuan-Hung Vu, Matthieu Cord, Patrick Pérez
While fully-supervised deep learning yields good models for urban scene semantic segmentation, these models struggle to generalize to new environments with different lighting or weather conditions for instance.
2 code implementations • ECCV 2020 • Arthur Douillard, Matthieu Cord, Charles Ollion, Thomas Robert, Eduardo Valle
Lifelong learning has attracted much attention, but existing works still struggle to fight catastrophic forgetting and accumulate knowledge over long stretches of incremental learning.
no code implementations • 14 Apr 2020 • Arnaud Dapogny, Kévin Bailly, Matthieu Cord
Head pose estimation and face alignment constitute a backbone preprocessing for many applications relying on face analysis.
no code implementations • 2 Apr 2020 • Maxime Bucher, Tuan-Hung Vu, Matthieu Cord, Patrick Pérez
In this work, we define and address a novel domain adaptation (DA) problem in semantic scene segmentation, where the target domain not only exhibits a data distribution shift w. r. t.
1 code implementation • CVPR 2020 • Spyros Gidaris, Andrei Bursuc, Nikos Komodakis, Patrick Pérez, Matthieu Cord
Inspired by the success of NLP methods in this area, in this work we propose a self-supervised approach based on spatially dense image descriptions that encode discrete visual concepts, here called visual words.
1 code implementation • ECCV 2020 • Himalaya Jain, Spyros Gidaris, Nikos Komodakis, Patrick Pérez, Matthieu Cord
Knowledge distillation refers to the process of training a compact student network to achieve better accuracy by learning from a high capacity teacher network.
1 code implementation • NeurIPS 2019 • Remi Cadene, Corentin Dancette, Hedi Ben Younes, Matthieu Cord, Devi Parikh
We propose RUBi, a new learning strategy to reduce biases in any VQA model.
1 code implementation • NeurIPS 2019 • Charles Corbière, Nicolas Thome, Avner Bar-Hen, Matthieu Cord, Patrick Pérez
In this paper, we propose a new target criterion for model confidence, corresponding to the True Class Probability (TCP).
no code implementations • 7 Nov 2019 • Victor Besnier, Himalaya Jain, Andrei Bursuc, Matthieu Cord, Patrick Pérez
This naturally brings the question: Can we train a classifier only on the generated data?
no code implementations • 15 Oct 2019 • Antoine Saporta, Yifu Chen, Michael Blot, Matthieu Cord
Studies on generalization performance of machine learning algorithms under the scope of information theory suggest that compressed representations can guarantee good generalization, inspiring many compression-based regularization methods.
1 code implementation • NeurIPS 2019 • Charles Corbière, Nicolas Thome, Avner Bar-Hen, Matthieu Cord, Patrick Pérez
In this paper, we propose a new target criterion for model confidence, corresponding to the True Class Probability (TCP).
no code implementations • NeurIPS 2019 • Daniel Brooks, Olivier Schwander, Frederic Barbaresco, Jean-Yves Schneider, Matthieu Cord
Covariance matrices have attracted attention for machine learning applications due to their capacity to capture interesting structure in the data.
1 code implementation • 24 Jun 2019 • Remi Cadene, Corentin Dancette, Hedi Ben-Younes, Matthieu Cord, Devi Parikh
We propose RUBi, a new learning strategy to reduce biases in any VQA model.
Ranked #7 on Visual Question Answering (VQA) on VQA-CP
1 code implementation • ICCV 2019 • Spyros Gidaris, Andrei Bursuc, Nikos Komodakis, Patrick Pérez, Matthieu Cord
Few-shot learning and self-supervised learning address different facets of the same problem: how to train a model with little or no labeled data.
1 code implementation • 3 Jun 2019 • Thomas Robert, Nicolas Thome, Matthieu Cord
To effectively separate the information, we propose to use a combination of regular and adversarial classifiers to guide the two branches in specializing for class and attribute information respectively.
2 code implementations • NeurIPS 2019 • Maxime Bucher, Tuan-Hung Vu, Matthieu Cord, Patrick Pérez
Semantic segmentation models are limited in their ability to scale to large numbers of object classes.
Ranked #1 on Zero-Shot Learning on PASCAL Context
no code implementations • 6 May 2019 • Yifu Chen, Arnaud Dapogny, Matthieu Cord
As a result, the predictions outputted by such networks usually struggle to accurately capture the object boundaries and exhibit holes inside the objects.
no code implementations • 6 May 2019 • Arnaud Dapogny, Matthieu Cord, Patrick Perez
Image completion is the problem of generating whole images from fragments only.
1 code implementation • CVPR 2019 • Martin Engilberge, Louis Chevallier, Patrick Pérez, Matthieu Cord
Our approach is based on a deep architecture that approximates the sorting of arbitrary sets of scores.
no code implementations • ICCV 2019 • Arnaud Dapogny, Kévin Bailly, Matthieu Cord
Face Alignment is an active computer vision domain, that consists in localizing a number of facial landmarks that vary across datasets.
Ranked #22 on Face Alignment on WFLW
2 code implementations • ICCV 2019 • Tuan-Hung Vu, Himalaya Jain, Maxime Bucher, Matthieu Cord, Patrick Pérez
As a result, the performance of the trained semantic segmentation model on the target domain is boosted.
Ranked #17 on Image-to-Image Translation on SYNTHIA-to-Cityscapes
1 code implementation • CVPR 2019 • Remi Cadene, Hedi Ben-Younes, Matthieu Cord, Nicolas Thome
In this paper, we propose MuRel, a multimodal relational network which is learned end-to-end to reason over real images.
Ranked #1 on Visual Question Answering (VQA) on TDIUC
1 code implementation • 31 Jan 2019 • Hedi Ben-Younes, Rémi Cadene, Nicolas Thome, Matthieu Cord
We demonstrate the practical interest of our fusion model by using BLOCK for two challenging tasks: Visual Question Answering (VQA) and Visual Relationship Detection (VRD), where we design end-to-end learnable architectures for representing relevant interactions between modalities.
1 code implementation • NeurIPS 2018 • Taylor Mordan, Nicolas Thome, Gilles Henaff, Matthieu Cord
Multi-Task Learning (MTL) is appealing for deep learning regularization.
4 code implementations • CVPR 2019 • Tuan-Hung Vu, Himalaya Jain, Maxime Bucher, Matthieu Cord, Patrick Pérez
Semantic segmentation is a key problem for many computer vision tasks.
Ranked #5 on Domain Adaptation on Panoptic SYNTHIA-to-Mapillary
no code implementations • ECCV 2018 • Thomas Robert, Nicolas Thome, Matthieu Cord
In this paper, we introduce a new model for leveraging unlabeled data to improve generalization performances of image classifiers: a two-branch encoder-decoder architecture called HybridNet.
Ranked #52 on Image Classification on STL-10
no code implementations • CVPR 2018 • Ãloi Mehr, André Lieutier, Fernando Sanchez Bermudez, Vincent Guitteny, Nicolas Thome, Matthieu Cord
Typically, we propose to quotient the space of 3D models by the action of rotations.
1 code implementation • 14 May 2018 • Michael Blot, Thomas Robert, Nicolas Thome, Matthieu Cord
Regularization is a big issue for training deep neural networks.
1 code implementation • 2 May 2018 • Micael Carvalho, Rémi Cadène, David Picard, Laure Soulier, Matthieu Cord
Recent advances in the machine learning community allowed different use cases to emerge, as its association to domains like cooking which created the computational cuisine.
1 code implementation • 30 Apr 2018 • Micael Carvalho, Rémi Cadène, David Picard, Laure Soulier, Nicolas Thome, Matthieu Cord
Designing powerful tools that support cooking activities has rapidly gained popularity due to the massive amounts of available data, as well as recent advances in machine learning that are capable of analyzing them.
Ranked #9 on Cross-Modal Retrieval on Recipe1M
no code implementations • 29 Apr 2018 • Michael Blot, Thomas Robert, Nicolas Thome, Matthieu Cord
Regularization is a big issue for training deep neural networks.
1 code implementation • CVPR 2018 • Martin Engilberge, Louis Chevallier, Patrick Pérez, Matthieu Cord
Several works have proposed to learn a two-path neural network that maps images and texts, respectively, to a same shared Euclidean space where geometry captures useful semantic relationships.
no code implementations • 4 Apr 2018 • Michael Blot, David Picard, Matthieu Cord
We address the issue of speeding up the training of convolutional neural networks by studying a distributed method adapted to stochastic gradient descent.
no code implementations • ICLR 2018 • Michael Blot, Thomas Robert, Nicolas Thome, Matthieu Cord
Regularization is a big issue for training deep neural networks.
no code implementations • 19 Jul 2017 • Taylor Mordan, Nicolas Thome, Matthieu Cord, Gilles Henaff
Existing region-based object detectors are limited to regions with fixed box geometry to represent objects, even if those are highly non-rectangular.
2 code implementations • CVPR 2017 • Thibaut Durand, Taylor Mordan, Nicolas Thome, Matthieu Cord
This paper introduces WILDCAT, a deep learning method which jointly aims at aligning image regions for gaining spatial invariance and learning strongly localized features.
Ranked #3 on Weakly Supervised Object Detection on MS COCO
6 code implementations • ICCV 2017 • Hedi Ben-Younes, Rémi Cadene, Matthieu Cord, Nicolas Thome
Bilinear models provide an appealing framework for mixing and merging information in Visual Question Answering (VQA) tasks.
Ranked #35 on Visual Question Answering (VQA) on VQA v2 test-std
1 code implementation • 29 Nov 2016 • Michael Blot, David Picard, Matthieu Cord, Nicolas Thome
We address the issue of speeding up the training of convolutional networks.
no code implementations • 25 Oct 2016 • Michael Blot, Matthieu Cord, Nicolas Thome
Convolutional neural networks (CNN) are widely used in computer vision, especially in image classification.
1 code implementation • 18 Oct 2016 • Rémi Cadène, Nicolas Thome, Matthieu Cord
Our last contribution is a framework, build on top of Torch7, for training and testing deep models on any visual recognition tasks and on datasets of any scale.
1 code implementation • 18 Oct 2016 • Rémi Cadène, Thomas Robert, Nicolas Thome, Matthieu Cord
Our approach is among the three best to tackle the M2CAI Workflow challenge.
1 code implementation • CVPR 2016 • Thibaut Durand, Nicolas Thome, Matthieu Cord
In this paper, we introduce a novel framework for WEakly supervised Learning of Deep cOnvolutional neural Networks (WELDON).
no code implementations • CVPR 2016 • Marc T. Law, Yao-Liang Yu, Matthieu Cord, Eric P. Xing
Clustering is the task of grouping a set of objects so that objects in the same cluster are more similar to each other than to those in other clusters.
1 code implementation • 11 May 2016 • Micael Carvalho, Matthieu Cord, Sandra Avila, Nicolas Thome, Eduardo Valle
In recent years, deep architectures have been used for transfer learning with state-of-the-art performance in many datasets.
no code implementations • ICCV 2015 • Thibaut Durand, Nicolas Thome, Matthieu Cord
For ranking, we propose efficient solutions to exactly solve the inference and the loss-augmented problems.
1 code implementation • IEEE 2015 • Xin Wang, Devinder Kumar, Nicolas Thome, Matthieu Cord, Frederic Precioso
We present deep experiments of recipe recognition on our dataset using visual, textual information and fusion.
no code implementations • CVPR 2014 • Marc T. Law, Nicolas Thome, Matthieu Cord
This paper introduces a regularization method to explicitly control the rank of a learned symmetric positive semidefinite distance matrix in distance metric learning.
no code implementations • 20 Dec 2013 • Gabriel Dulac-Arnold, Ludovic Denoyer, Nicolas Thome, Matthieu Cord, Patrick Gallinari
In this paper, we investigate a new framework for image classification that adaptively generates spatial representations.
no code implementations • NeurIPS 2013 • Hanlin Goh, Nicolas Thome, Matthieu Cord, Joo-Hwee Lim
We suggest a deep learning strategy that bridges the gap between the two phases, resulting in a three-phase learning procedure.
no code implementations • CVPR 2013 • Christian Theriault, Nicolas Thome, Matthieu Cord
In this paper, we address the challenging problem of categorizing video sequences composed of dynamic natural scenes.