Publications (21)
CHMv2: Improvements in Global Canopy Height Mapping using DINOv3
John Brandt, Seungeun Yi, Jamie Tolan +9
Accurate canopy height information is essential for quantifying forest carbon, monitoring restoration and degradation, and assessing habitat structure, yet high-fidelity measuremen…
Disentangling the Factors of Convergence between Brains and Computer Vision Models
Joséphine Raugel, Marc Szafraniec, Huy V. Vo +5
Many AI models trained on natural images develop representations that resemble those of the human brain. However, the factors that drive this brain-model similarity remain poorly u…
Accelerating Transformer Inference and Training with 2:4 Activation Sparsity
Daniel Haziza, Timothy Chou, Dhruv Choudhary +7
In this paper, we demonstrate how to leverage 2:4 sparsity, a popular hardware-accelerated GPU sparsity pattern, to activations to accelerate large language model training and infe…
Efficient Universal Perception Encoder
Chenchen Zhu, Saksham Suri, Cijo Jose +8
Running AI models on smart edge devices can unlock versatile user experiences, but presents challenges due to limited compute and the need to handle multiple tasks simultaneously.…
VGGT-
Jianyuan Wang, Minghao Chen, Shangzhan Zhang +7
Recent feed-forward reconstruction models, such as VGGT, have proven competitive with traditional optimization-based reconstructors while also providing geometry-aware features use…
Low Bandwidth Video-Chat Compression using Deep Generative Models
Maxime Oquab, Pierre Stock, Oran Gafni +9
To unlock video chat for hundreds of millions of people hindered by poor connectivity or unaffordable data costs, we propose to authentically reconstruct faces on the receiver's de…
DensePose 3D: Lifting Canonical Surface Maps of Articulated Objects to the Third Dimension
Roman Shapovalov, David Novotny, Benjamin Graham +2
We tackle the problem of monocular 3D reconstruction of articulated objects like humans and animals. We contribute DensePose 3D, a method that can learn such reconstructions in a w…
Discovering Relationships between Object Categories via Universal Canonical Maps
Natalia Neverova, Artsiom Sanakoyeu, Patrick Labatut +2
We tackle the problem of learning the geometry of multiple categories of deformable objects jointly. Recent work has shown that it is possible to learn a unified dense pose predict…
AdaDINO: Context-Adaptive DINO-Distilled Vision Foundation Models for Efficient Open-Vocabulary Edge Inference
Yiwei Zhao, Yi Zheng, Huapeng Su +9
Always-on contextual AI runs language-aligned vision foundation models (VFMs) on edge devices, where the on-device model is the dominant continuous compute cost under strict latenc…
Back to the Features: DINO as a Foundation for Video World Models
Federico Baldassarre, Marc Szafraniec, Basile Terver +6
We present DINO-world, a powerful generalist video world model trained to predict future frames in the latent space of DINOv2. By leveraging a pre-trained image encoder and trainin…
Unsupervised Learning of 3D Object Categories from Videos in the Wild
Philipp Henzler, Jeremy Reizenstein, Patrick Labatut +4
Our goal is to learn a deep network that, given a small number of images of an object of a given category, reconstructs it in 3D. While several recent works have obtained analogous…
Misalignment Between Backpropagation and the Hierarchy of Brain Responses to Images
Joséphine Raugel, Maximilian Seitzer, Marc Szafraniec +6
Backpropagation is the core learning mechanism underlying deep learning. However, whether and how this algorithm is implemented in the brain remains highly debated. In particular,…
DINOv3
Oriane Siméoni, Huy V. Vo, Maximilian Seitzer +23
Self-supervised learning holds the promise of eliminating the need for manual data annotation, enabling models to scale effortlessly to massive datasets and larger architectures. B…
Automatic Data Curation for Self-Supervised Learning: A Clustering-Based Approach
Huy V. Vo, Vasil Khalidov, Timothée Darcet +10
Self-supervised features are the cornerstone of modern machine learning systems. They are typically pre-trained on data collections whose construction and curation typically requir…
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…
DINOv2 Meets Text: A Unified Framework for Image- and Pixel-Level Vision-Language Alignment
Cijo Jose, Théo Moutakanni, Dahyun Kang +11
Self-supervised visual foundation models produce powerful embeddings that achieve remarkable performance on a wide range of downstream tasks. However, unlike vision-language models…
Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction
Jeremy Reizenstein, Roman Shapovalov, Philipp Henzler +3
Traditional approaches for learning 3D object categories have been predominantly trained and evaluated on synthetic datasets due to the unavailability of real 3D-annotated category…
NeuroMorph: Unsupervised Shape Interpolation and Correspondence in One Go
Marvin Eisenberger, David Novotny, Gael Kerchenbaum +4
We present NeuroMorph, a new neural network architecture that takes as input two 3D shapes and produces in one go, i.e. in a single feed forward pass, a smooth interpolation and po…
Code Translation with Compiler Representations
Marc Szafraniec, Baptiste Roziere, Hugh Leather +3
In this paper, we leverage low-level compiler intermediate representations (IR) to improve code translation. Traditional transpilers rely on syntactic information and handcrafted r…
Continuous Surface Embeddings
Natalia Neverova, David Novotny, Vasil Khalidov +3
In this work, we focus on the task of learning and representing dense correspondences in deformable object categories. While this problem has been considered before, solutions so f…
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…