no code implementations • 22 Apr 2024 • Sophia Sirko-Galouchenko, Alexandre Boulch, Spyros Gidaris, Andrei Bursuc, Antonin Vobecky, Patrick Pérez, Renaud Marlet
Models pretrained with our method exhibit improved BEV semantic segmentation performance, particularly in low-data scenarios.
no code implementations • NeurIPS 2023 • Antonin Vobecky, Oriane Siméoni, David Hurych, Spyros Gidaris, Andrei Bursuc, Patrick Pérez, Josef Sivic
We describe an approach to predict open-vocabulary 3D semantic voxel occupancy map from input 2D images with the objective of enabling 3D grounding, segmentation and retrieval of free-form language queries.
3D Semantic Occupancy Prediction 3D Semantic Segmentation +3
no code implementations • 23 Dec 2023 • Gianni Franchi, Olivier Laurent, Maxence Leguéry, Andrei Bursuc, Andrea Pilzer, Angela Yao
Deep Neural Networks (DNNs) are powerful tools for various computer vision tasks, yet they often struggle with reliable uncertainty quantification - a critical requirement for real-world applications.
1 code implementation • 19 Dec 2023 • Monika Wysoczańska, Oriane Siméoni, Michaël Ramamonjisoa, Andrei Bursuc, Tomasz Trzciński, Patrick Pérez
We propose to locally improve dense MaskCLIP features, which are computed with a simple modification of CLIP's last pooling layer, by integrating localization priors extracted from self-supervised features.
1 code implementation • 29 Nov 2023 • Mohammad Fahes, Tuan-Hung Vu, Andrei Bursuc, Patrick Pérez, Raoul de Charette
Generalization to new domains not seen during training is one of the long-standing challenges in deploying neural networks in real-world applications.
1 code implementation • 26 Oct 2023 • Gilles Puy, Spyros Gidaris, Alexandre Boulch, Oriane Siméoni, Corentin Sautier, Patrick Pérez, Andrei Bursuc, Renaud Marlet
In particular, thanks to our scalable distillation method named ScaLR, we show that scaling the 2D and 3D backbones and pretraining on diverse datasets leads to a substantial improvement of the feature quality.
no code implementations • 27 Sep 2023 • Xuanlong Yu, Yi Zuo, Zitao Wang, Xiaowen Zhang, Jiaxuan Zhao, Yuting Yang, Licheng Jiao, Rui Peng, Xinyi Wang, Junpei Zhang, Kexin Zhang, Fang Liu, Roberto Alcover-Couso, Juan C. SanMiguel, Marcos Escudero-Viñolo, Hanlin Tian, Kenta Matsui, Tianhao Wang, Fahmy Adan, Zhitong Gao, Xuming He, Quentin Bouniot, Hossein Moghaddam, Shyam Nandan Rai, Fabio Cermelli, Carlo Masone, Andrea Pilzer, Elisa Ricci, Andrei Bursuc, Arno Solin, Martin Trapp, Rui Li, Angela Yao, Wenlong Chen, Ivor Simpson, Neill D. F. Campbell, Gianni Franchi
This paper outlines the winning solutions employed in addressing the MUAD uncertainty quantification challenge held at ICCV 2023.
no code implementations • 19 Sep 2023 • Clement Laroudie, Andrei Bursuc, Mai Lan Ha, Gianni Franchi
This paper examines the robustness of a multi-modal computer vision model, CLIP (Contrastive Language-Image Pretraining), in the context of unsupervised learning.
no code implementations • 1 Aug 2023 • Marwane Hariat, Olivier Laurent, Rémi Kazmierczak, Shihao Zhang, Andrei Bursuc, Angela Yao, Gianni Franchi
We propose a novel approach to improve the robustness of semantic segmentation techniques by leveraging the synergy between label-to-image generators and image-to-label segmentation models.
no code implementations • 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.
1 code implementation • CVPR 2023 • Angelika Ando, Spyros Gidaris, Andrei Bursuc, Gilles Puy, Alexandre Boulch, Renaud Marlet
(c) We refine pixel-wise predictions with a convolutional decoder and a skip connection from the convolutional stem to combine low-level but fine-grained features of the the convolutional stem with the high-level but coarse predictions of the ViT encoder.
1 code implementation • ICCV 2023 • Mohammad Fahes, Tuan-Hung Vu, Andrei Bursuc, Patrick Pérez, Raoul de Charette
In this paper, we propose the task of 'Prompt-driven Zero-shot Domain Adaptation', where we adapt a model trained on a source domain using only a general description in natural language of the target domain, i. e., a prompt.
1 code implementation • 6 Dec 2022 • Mohammad Fahes, Tuan-Hung Vu, Andrei Bursuc, Patrick Pérez, Raoul de Charette
In this paper, we propose the task of `Prompt-driven Zero-shot Domain Adaptation', where we adapt a model trained on a source domain using only a general description in natural language of the target domain, i. e., a prompt.
1 code implementation • 17 Oct 2022 • Olivier Laurent, Adrien Lafage, Enzo Tartaglione, Geoffrey Daniel, Jean-Marc Martinez, Andrei Bursuc, Gianni Franchi
Deep Ensembles (DE) are a prominent approach for achieving excellent performance on key metrics such as accuracy, calibration, uncertainty estimation, and out-of-distribution detection.
1 code implementation • 25 Jul 2022 • Huy V. Vo, Oriane Siméoni, Spyros Gidaris, Andrei Bursuc, Patrick Pérez, Jean Ponce
On COCO, using on average 10 fully-annotated images per class, or equivalently 1% of the training set, BiB also reduces the performance gap (in AP) between the weakly-supervised detector and the fully-supervised Fast RCNN by over 70%, showing a good trade-off between performance and data efficiency.
no code implementations • 22 Jul 2022 • Timo Sämann, Ahmed Mostafa Hammam, Andrei Bursuc, Christoph Stiller, Horst-Michael Groß
Albeit effective, only few works haveimproved the understanding and the performance of weight averaging. Here, we revisit this approach and show that a simple weight fusion (WF)strategy can lead to a significantly improved predictive performance andcalibration.
1 code implementation • 20 Jul 2022 • Gianni Franchi, Xuanlong Yu, Andrei Bursuc, Emanuel Aldea, Severine Dubuisson, David Filliat
Predictive uncertainty estimation is essential for deploying Deep Neural Networks in real-world autonomous systems.
no code implementations • 18 Jul 2022 • Victor Besnier, Andrei Bursuc, David Picard, Alexandre Briot
To address this issue, we build upon the recent ObsNet approach by providing object instance knowledge to the observer.
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
1 code implementation • CVPR 2022 • Corentin Sautier, Gilles Puy, Spyros Gidaris, Alexandre Boulch, Andrei Bursuc, Renaud Marlet
In this context, we propose a self-supervised pre-training method for 3D perception models that is tailored to autonomous driving data.
no code implementations • 29 Mar 2022 • Subhrajyoti Dasgupta, Arindam Das, Senthil Yogamani, Sudip Das, Ciaran Eising, Andrei Bursuc, Ujjwal Bhattacharya
Shadows are frequently encountered natural phenomena that significantly hinder the performance of computer vision perception systems in practical settings, e. g., autonomous driving.
no code implementations • 28 Mar 2022 • Kevin Osanlou, Jeremy Frank, Andrei Bursuc, Tristan Cazenave, Eric Jacopin, Christophe Guettier, J. Benton
Moreover, we leverage a graph neural network as a heuristic for tree search guidance.
1 code implementation • 23 Mar 2022 • Ioannis Kakogeorgiou, Spyros Gidaris, Bill Psomas, Yannis Avrithis, Andrei Bursuc, Konstantinos Karantzalos, Nikos Komodakis
In this work, we argue that image token masking differs from token masking in text, due to the amount and correlation of tokens in an image.
1 code implementation • 21 Mar 2022 • Antonin Vobecky, David Hurych, Oriane Siméoni, Spyros Gidaris, Andrei Bursuc, Patrick Pérez, Josef Sivic
This work investigates learning pixel-wise semantic image segmentation in urban scenes without any manual annotation, just from the raw non-curated data collected by cars which, equipped with cameras and LiDAR sensors, drive around a city.
3 code implementations • 2 Mar 2022 • Gianni Franchi, Xuanlong Yu, Andrei Bursuc, Angel Tena, Rémi Kazmierczak, Séverine Dubuisson, Emanuel Aldea, David Filliat
However, disentangling the different types and sources of uncertainty is non trivial for most datasets, especially since there is no ground truth for uncertainty.
2 code implementations • 29 Sep 2021 • Oriane Siméoni, Gilles Puy, Huy V. Vo, Simon Roburin, Spyros Gidaris, Andrei Bursuc, Patrick Pérez, Renaud Marlet, Jean Ponce
We also show that training a class-agnostic detector on the discovered objects boosts results by another 7 points.
Ranked #4 on Weakly-Supervised Object Localization on CUB-200-2011 (Top-1 Localization Accuracy metric)
1 code implementation • ICCV 2021 • Victor Besnier, Andrei Bursuc, David Picard, Alexandre Briot
In this paper, we tackle the detection of out-of-distribution (OOD) objects in semantic segmentation.
Out-of-Distribution Detection Out of Distribution (OOD) Detection +2
no code implementations • 2 Aug 2021 • Kevin Osanlou, Christophe Guettier, Andrei Bursuc, Tristan Cazenave, Eric Jacopin
The uncertain criterion represents the feasibility of driving through the path without requiring human intervention.
no code implementations • 2 Aug 2021 • Kevin Osanlou, Andrei Bursuc, Christophe Guettier, Tristan Cazenave, Eric Jacopin
More specifically, a graph neural network is used to assist the branch and bound algorithm in handling constraints associated with a desired solution path.
1 code implementation • 2 Aug 2021 • Gianni Franchi, Nacim Belkhir, Mai Lan Ha, Yufei Hu, Andrei Bursuc, Volker Blanz, Angela Yao
Along with predictive performance and runtime speed, reliability is a key requirement for real-world semantic segmentation.
no code implementations • 2 Aug 2021 • Kevin Osanlou, Christophe Guettier, Andrei Bursuc, Tristan Cazenave, Eric Jacopin
In this paper, we focus on shortest path search with mandatory nodes on a given connected graph.
no code implementations • 2 Aug 2021 • Kevin Osanlou, Jeremy Frank, J. Benton, Andrei Bursuc, Christophe Guettier, Eric Jacopin, Tristan Cazenave
Scheduling in the presence of uncertainty is an area of interest in artificial intelligence due to the large number of applications.
no code implementations • 25 Mar 2021 • Julien Rebut, Andrei Bursuc, Patrick Pérez
Robustness to various image corruptions, caused by changing weather conditions or sensor degradation and aging, is crucial for safety when such vehicles are deployed in the real world.
2 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)
2 code implementations • 4 Dec 2020 • Gianni Franchi, Andrei Bursuc, Emanuel Aldea, Severine Dubuisson, Isabelle Bloch
Bayesian neural networks (BNNs) have been long considered an ideal, yet unscalable solution for improving the robustness and the predictive uncertainty of deep neural networks.
Ranked #132 on Image Classification on CIFAR-10
1 code implementation • 23 Jun 2020 • Simon Roburin, Yann de Mont-Marin, Andrei Bursuc, Renaud Marlet, Patrick Pérez, Mathieu Aubry
Normalization Layers (NLs) are widely used in modern deep-learning architectures.
no code implementations • 1 Jun 2020 • Gianni Franchi, Andrei Bursuc, Emanuel Aldea, Severine Dubuisson, Isabelle Bloch
This is due to the fact that modern DNNs are usually uncalibrated and we cannot characterize their epistemic uncertainty.
no code implementations • 9 Mar 2020 • Thibault Buhet, Emilie Wirbel, Andrei Bursuc, Xavier Perrotton
Our model processes ego vehicle front-facing camera images and bird-eye view grid, computed from Lidar point clouds, with detections of past and present objects, in order to generate multiple trajectories for both ego vehicle and its neighbors.
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.
no code implementations • 7 Jan 2020 • Isabelle Leang, Ganesh Sistu, Fabian Burger, Andrei Bursuc, Senthil Yogamani
Deep multi-task networks are of particular interest for autonomous driving systems.
no code implementations • ECCV 2020 • Gianni Franchi, Andrei Bursuc, Emanuel Aldea, Severine Dubuisson, Isabelle Bloch
During training, the weights of a Deep Neural Network (DNN) are optimized from a random initialization towards a nearly optimum value minimizing a loss function.
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?
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.
no code implementations • CVPR 2019 • Yann Lifchitz, Yannis Avrithis, Sylvaine Picard, Andrei Bursuc
Training deep neural networks from few examples is a highly challenging and key problem for many computer vision tasks.
3 code implementations • 27 Nov 2018 • Arun Mukundan, Giorgos Tolias, Andrei Bursuc, Hervé Jégou, Ondřej Chum
We propose a multiple-kernel local-patch descriptor based on efficient match kernels from pixel gradients.