Bookmarks

3D Gaussian Splatting

V-JEPA: Revisiting Feature Prediction for Learning Visual Representations from Video (Explained)

Paper walk-through of V-JEPA, detailing a predictive video representation model trained without labels for downstream vision tasks.

How fly neurons compute the direction of visual motion

Gemini: A Family of Highly Capable Multimodal Models

MotionGPT: Human Motion as a Foreign Language

Subcategories