[CVPR 2025 Highlight] Video Depth Anything: Consistent Depth Estimation for Super-Long Videos

A Large-scale Video Action Dataset

Official implementation of SimToolReal: An Object-Centric Policy for Zero-Shot Dexterous Tool Manipulation

Code for "SceneSmith: Agentic Generation of Simulation-Ready Indoor Scenes", ICML 2026 Spotlight

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

Puffing up reinforcement learning

Official implementation of Video-DPM

Reference PyTorch implementation and models for DINOv3

🍽️ Annotations for the public release of the EPIC-KITCHENS-100 dataset

Large-scale text-video dataset. 10 million captioned short videos.

WiLoR: End-to-end 3D hand localization and reconstruction in-the-wild

DEXTRAH

phone teleoperation for robots

UMI-on-Air: Embodiment-Aware Guidance for Embodiment-Agnostic Visuomotor Policies

code for the paper Predicting Point Tracks from Internet Videos enables Diverse Zero-Shot Manipulation

Official codebase for "Any-point Trajectory Modeling for Policy Learning"

Generate synthetic GelSight tactile images and depth maps from 3D meshes using Blender.

Simplifying diffusion/flow policies by treating action trajectories as flow trajectories

Streaming Diffusion Policy: Fast Policy Synthesis with Variable Noise Diffusion Models

Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots

Official implementation of Get a Grip: Multi-Finger Grasp Evaluation at Scale Enables Robust Sim-to-Real Transfer

(RA-L 2025) UniT: Data Efficient Tactile Representation with Generalization to Unseen Objects

world modeling challenge for humanoid robots

This is the official codebase for the paper "Sensor-Invariant Tactile Representation" (ICLR 2025).

GelSight SDK (including calibration tools)