Explanipedia

3D Human Pose and Shape Estimation from LiDAR Point Clouds: A Review Open

Salma Galaaoui, Eduardo Valle, David Picard, Nermin Samet · 2025

In this paper, we present a comprehensive review of 3D human pose estimation and human mesh recovery from in-the-wild LiDAR point clouds. We compare existing approaches across several key dimensions, and propose a structured taxonomy to cl…

Closing the Performance Gap in Generative Recommenders with Collaborative Tokenization and Efficient Modeling Open

Simon Lepage, Jérémie Mary, David Picard · 2025

Recent work has explored generative recommender systems as an alternative to traditional ID-based models, reframing item recommendation as a sequence generation task over discrete item tokens. While promising, such methods often underperfo…

Around the World in 80 Timesteps: A Generative Approach to Global Visual Geolocation Open

Nicolas Dufour, David Picard, Vicky Kalogeiton, Loïc Landrieu · 2024

Global visual geolocation predicts where an image was captured on Earth. Since images vary in how precisely they can be localized, this task inherently involves a significant degree of ambiguity. However, existing approaches are determinis…

PoM: Efficient Image and Video Generation with the Polynomial Mixer Open

David Picard, Nicolas Dufour · 2024

Diffusion models based on Multi-Head Attention (MHA) have become ubiquitous to generate high quality images and videos. However, encoding an image or a video as a sequence of patches results in costly attention patterns, as the requirement…

PAFUSE: Part-based Diffusion for 3D Whole-Body Pose Estimation Open

Nermin Samet, Cédric Rommel, David Picard, Eduardo Valle · 2024

We introduce a novel approach for 3D whole-body pose estimation, addressing the challenge of scale -- and deformability -- variance across body parts brought by the challenge of extending the 17 major joints on the human body to fine-grain…

Don't drop your samples! Coherence-aware training benefits Conditional diffusion Open

Nicolas Dufour, Victor Besnier, Vicky Kalogeiton, David Picard · 2024

Conditional diffusion models are powerful generative models that can leverage various types of conditional information, such as class labels, segmentation masks, or text captions. However, in many real-world scenarios, conditional informat…

Analysis of Classifier-Free Guidance Weight Schedulers Open

Xi Wang, Nicolas Dufour, Nefeli Andreou, Marie-Paule Cani, Victoria Abrevaya , et al. · 2024

Classifier-Free Guidance (CFG) enhances the quality and condition adherence of text-to-image diffusion models. It operates by combining the conditional and unconditional predictions using a fixed weight. However, recent works vary the weig…

Multiple Locally Linear Kernel Machines Open

David Picard · 2024

In this paper we propose a new non-linear classifier based on a combination of locally linear classifiers. A well known optimization formulation is given as we cast the problem in a $\ell_1$ Multiple Kernel Learning (MKL) problem using man…

An Analysis of Initial Training Strategies for Exemplar-Free Class-Incremental Learning Open

Grégoire Petit, Michaël Soumm, Eva Feillet, Adrian Popescu, Bertrand Delezoide , et al. · 2024

Class-Incremental Learning (CIL) aims to build classification models from data streams. At each step of the CIL process, new classes must be integrated into the model. Due to catastrophic forgetting, CIL is particularly challenging when ex…

Toward Unsupervised Visual Reasoning: Do Off-the-Shelf Features Know How to Reason? Open

Monika Wysoczańska, Tom Monnier, T. P. Trzcinski, David Picard · 2024

Recent advances in visual representation learning allowed for the construction of a plethora of powerful features that are ready to use for numerous downstream tasks. Contrary to existing representation evaluations typically based on image…

Image Compression using only Attention based Neural Networks Open

Natacha Luka, Romain Negrel, David Picard · 2023

In recent research, Learned Image Compression has gained prominence for its capacity to outperform traditional handcrafted pipelines, especially at low bit-rates. While existing methods incorporate convolutional priors with occasional atte…

Contraction d'une image en une petite séquence de tokens avec des modules d'attention croisée Open

Natacha Luka, Romain Negrel, David Picard · 2023

International audience

An Analysis of Initial Training Strategies for Exemplar-Free Class-Incremental Learning Open

Grégoire Petit, Michaël Soumm, Eva Feillet, Adrian Popescu, Bertrand Delezoide , et al. · 2023

Class-Incremental Learning (CIL) aims to build classification models from data streams. At each step of the CIL process, new classes must be integrated into the model. Due to catastrophic forgetting, CIL is particularly challenging when ex…

LRVS-Fashion: Extending Visual Search with Referring Instructions Open

Simon Lepage, Jérémie Mary, David Picard · 2023

This paper introduces a new challenge for image similarity search in the context of fashion, addressing the inherent ambiguity in this domain stemming from complex images. We present Referred Visual Search (RVS), a task allowing users to d…

SSP-Net: Scalable sequential pyramid networks for real-Time 3D human pose regression Open

Diogo Luvizon, Hedi Tabia, David Picard · 2023

Alphazzle: Jigsaw Puzzle Solver with Deep Monte-Carlo Tree Search Open

Marie-Morgane Paumard, Hedi Tabia, David Picard · 2023

Solving jigsaw puzzles requires to grasp the visual features of a sequence of patches and to explore efficiently a solution space that grows exponentially with the sequence length. Therefore, visual deep reinforcement learning (DRL) should…

FeTrIL: Feature Translation for Exemplar-Free Class-Incremental Learning Open

Grégoire Petit, Adrian Popescu, Hugo Schindler, David Picard · 2023

Exemplar-free class-incremental learning is very challenging due to the negative effect of catastrophic forgetting. A balance between stability and plasticity of the incremental process is needed in order to obtain good accuracy for past a…

FeTrIL: Feature Translation for Exemplar-Free Class-Incremental Learning Open

Grégoire Petit, Adrian Popescu, Hugo Schindler, David Picard, Bertrand Delezoide · 2023

International audience

Towards Unsupervised Visual Reasoning: Do Off-The-Shelf Features Know How to Reason? Open

Monika Wysoczańska, Tom Monnier, T. P. Trzcinski, David Picard · 2022

Recent advances in visual representation learning allowed to build an abundance of powerful off-the-shelf features that are ready-to-use for numerous downstream tasks. This work aims to assess how well these features preserve information a…

H3WB: Human3.6M 3D WholeBody Dataset and Benchmark Open

Yue Zhu, Nermin Samet, David Picard · 2022

We present a benchmark for 3D human whole-body pose estimation, which involves identifying accurate 3D keypoints on the entire human body, including face, hands, body, and feet. Currently, the lack of a fully annotated and accurate 3D whol…

FeTrIL: Feature Translation for Exemplar-Free Class-Incremental Learning Open

Grégoire Petit, Adrian Popescu, Hugo Schindler, David Picard, Bertrand Delezoide · 2022

Exemplar-free class-incremental learning is very challenging due to the negative effect of catastrophic forgetting. A balance between stability and plasticity of the incremental process is needed in order to obtain good accuracy for past a…

Improving Deep Metric Learning with Virtual Classes and Examples Mining Open

Pierre Jacob, David Picard, Aymeric Histace · 2022

International audience

SCAM! Transferring humans between images with Semantic Cross Attention Modulation Open

Nicolas Dufour, David Picard, Vicky Kalogeiton · 2022

A large body of recent work targets semantically conditioned image generation. Most such methods focus on the narrower task of pose transfer and ignore the more challenging task of subject transfer that consists in not only transferring th…

Decanus to Legatus: Synthetic training for 2D-3D human pose lifting Open

Yue Zhu, David Picard · 2022

3D human pose estimation is a challenging task because of the difficulty to acquire ground-truth data outside of controlled environments. A number of further issues have been hindering progress in building a universal and robust model for …

PlaStIL: Plastic and Stable Memory-Free Class-Incremental Learning Open

Grégoire Petit, Adrian Popescu, Eden Belouadah, David Picard, Bertrand Delezoide · 2022

Plasticity and stability are needed in class-incremental learning in order to learn from new data while preserving past knowledge. Due to catastrophic forgetting, finding a compromise between these two properties is particularly challengin…

Achievable Rates and Machine Learning Detection of Faster than Nyquist Spectrally Efficient FDM Open

Arsenia Chorti, David Picard · 2022

International audience

Unveiling the Latent Space Geometry of Push-Forward Generative Models Open

Thibaut Issenhuth, Ugo Tanielian, Jérémie Mary, David Picard · 2022

Many deep generative models are defined as a push-forward of a Gaussian measure by a continuous generator, such as Generative Adversarial Networks (GANs) or Variational Auto-Encoders (VAEs). This work explores the latent space of such deep…

Instance-Aware Observer Network for Out-of-Distribution Object Segmentation Open

Victor Besnier, Andrei Bursuc, David Picard, Alexandre Briot · 2022

Recent works on predictive uncertainty estimation have shown promising results on Out-Of-Distribution (OOD) detection for semantic segmentation. However, these methods struggle to precisely locate the point of interest in the image, i.e, t…

Consensus-Based Optimization for 3D Human Pose Estimation in Camera Coordinates Open

Diogo Luvizon, David Picard, Hedi Tabia · 2022

SCAM! Transferring Humans Between Images with Semantic Cross Attention Modulation Open

Nicolas Dufour, David Picard, Vicky Kalogeiton · 2022

David Picard YOU? Author Swipe