1.1k citations · 1.4k across the 16 of their papers we have counts for
18 papers
EV-JEPA: Elastic V-JEPA for Any Device and Cross-Device Inference
Anxhelo Diko, Jieyu Lin, Stefano Ambrogio +5
Foundation models like V-JEPA are increasingly deployed across a heterogeneous edge-cloud continuum spanning smart glasses, phones, robots, and servers. Yet, they are typically tra…
Global Exponential Convergence of Two-Layer Linear Network Training
Stephen Y Zhang, Gabriel Peyré
We prove global exponential (linear) convergence with an explicit rate in the rich scaling for wide two-layer linear networks trained with smooth Polyak-Lojasiewicz predictor losse…
V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
Mido Assran, Adrien Bardes, David Fan +27
A major challenge for modern AI is to learn to understand the world and learn to act largely by observation. This paper explores a self-supervised approach that combines internet-s…
Intuitive physics understanding emerges from self-supervised pretraining on natural videos
Quentin Garrido, Nicolas Ballas, Mahmoud Assran +5
We investigate the emergence of intuitive physics understanding in general-purpose deep neural network models trained to predict masked regions in natural videos. Leveraging the vi…
Modeling Caption Diversity in Contrastive Vision-Language Pretraining
Samuel Lavoie, Polina Kirichenko, Mark Ibrahim +4
There are a thousand ways to caption an image. Contrastive Language Pretraining (CLIP) on the other hand, works by mapping an image and its caption to a single vector -- limiting h…
Learning and Leveraging World Models in Visual Representation Learning
Quentin Garrido, Mahmoud Assran, Nicolas Ballas +3
Joint-Embedding Predictive Architecture (JEPA) has emerged as a promising self-supervised approach that learns by leveraging a world model. While previously limited to predicting m…
Stochastic positional embeddings improve masked image modeling
Amir Bar, Florian Bordes, Assaf Shocher +6
Masked Image Modeling (MIM) is a promising self-supervised learning approach that enables learning from unlabeled images. Despite its recent success, learning good representations…
DINOv2: Learning Robust Visual Features without Supervision
Maxime Oquab, Timothée Darcet, Théo Moutakanni +23
The recent breakthroughs in natural language processing for model pretraining on large quantities of data have opened the way for similar foundation models in computer vision. Thes…
Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture
Mahmoud Assran, Quentin Duval, Ishan Misra +5
This paper demonstrates an approach for learning highly semantic image representations without relying on hand-crafted data-augmentations. We introduce the Image-based Joint-Embedd…
The Hidden Uniform Cluster Prior in Self-Supervised Learning
Mahmoud Assran, Randall Balestriero, Quentin Duval +6
A successful paradigm in representation learning is to perform self-supervised pretraining using tasks based on mini-batch statistics (e.g., SimCLR, VICReg, SwAV, MSN). We show tha…
Masked Siamese Networks for Label-Efficient Learning
Mahmoud Assran, Mathilde Caron, Ishan Misra +6
We propose Masked Siamese Networks (MSN), a self-supervised learning framework for learning image representations. Our approach matches the representation of an image view containi…
Semi-Supervised Learning of Visual Features by Non-Parametrically Predicting View Assignments with Support Samples
Mahmoud Assran, Mathilde Caron, Ishan Misra +4
This paper proposes a novel method of learning by predicting view assignments with support samples (PAWS). The method trains a model to minimize a consistency loss, which ensures t…
A Closer Look at Codistillation for Distributed Training
Shagun Sodhani, Olivier Delalleau, Mahmoud Assran +3
Codistillation has been proposed as a mechanism to share knowledge among concurrently trained models by encouraging them to represent the same function through an auxiliary loss. T…
Supervision Accelerates Pre-training in Contrastive Semi-Supervised Learning of Visual Representations
Mahmoud Assran, Nicolas Ballas, Lluis Castrejon +1
We investigate a strategy for improving the efficiency of contrastive learning of visual representations by leveraging a small amount of supervised information during pre-training.…
On the Convergence of Nesterov's Accelerated Gradient Method in Stochastic Settings
Mahmoud Assran, Michael Rabbat
We study Nesterov's accelerated gradient method with constant step-size and momentum parameters in the stochastic approximation setting (unbiased gradients with bounded variance) a…
Gossip-based Actor-Learner Architectures for Deep Reinforcement Learning
Mahmoud Assran, Joshua Romoff, Nicolas Ballas +2
Multi-simulator training has contributed to the recent success of Deep Reinforcement Learning by stabilizing learning and allowing for higher training throughputs. We propose Gossi…
Stochastic Gradient Push for Distributed Deep Learning
Mahmoud Assran, Nicolas Loizou, Nicolas Ballas +1
Distributed data-parallel algorithms aim to accelerate the training of deep neural networks by parallelizing the computation of large mini-batch gradient updates across multiple no…
Asynchronous Gradient-Push
Mahmoud Assran, Michael Rabbat
We consider a multi-agent framework for distributed optimization where each agent has access to a local smooth strongly convex function, and the collective goal is to achieve conse…