Diane Larlus

21

Papers

30

Total Citations

Papers (21)

UNIC: Universal Classification Models via Multi-teacher Distillation

LUDVIG: Learning-Free Uplifting of 2D Visual Features to Gaussian Splatting Scenes

Geo4D: Leveraging Video Generators for Geometric 4D Scene Reconstruction

Fisher Vectors Meet Neural Networks: A Hybrid Classification Architecture

AnchorNet: A Weakly Supervised Network to Learn Geometry-Sensitive Features for Semantic Matching

Beyond Instance-Level Image Retrieval: Leveraging Captions to Learn a Global Visual Representation for Semantic Retrieval

Self-Supervised Learning of Geometrically Stable Features Through Probabilistic Introspection

Continual Adaptation of Visual Representations via Domain Randomization and Meta-Learning

Probabilistic Embeddings for Cross-Modal Retrieval

On the Road to Online Adaptation for Semantic Image Segmentation

SLACK: Stable Learning of Augmentations With Cold-Start and KL Regularization

Fake It Till You Make It: Learning Transferable Representations From Synthetic ImageNet Clones

Learning 3D Object Categories by Looking Around Them

Fine-Grained Action Retrieval Through Multiple Parts-of-Speech Embeddings

Concept Generalization in Visual Representation Learning

Learning Visual Representations with Caption Annotations

Granularity-Aware Adaptation for Image Retrieval over Multiple Tasks

Layered Motion Fusion: Lifting Motion Segmentation to 3D in Egocentric Videos

DUNE: Distilling a Universal Encoder from Heterogeneous 2D and 3D Teachers

Hard Negative Mixing for Contrastive Learning

EPIC Fields: Marrying 3D Geometry and Video Understanding