ReKep: Spatio-Temporal Reasoning of Relational Keypoint Constraints for Robotic Manipulation

[ICLR 2025🎉] This is the official implementation of paper "Robots Pre-Train Robots: Manipulation-Centric Robotic Representation from Large-Scale Robot Datasets".

Awesome World Model for Robotics Papers

[CVPR 2024] Alpha-CLIP: A CLIP Model Focusing on Wherever You Want

A learning rate range test implementation in PyTorch

[CVPR 2022] DenseCLIP: Language-Guided Dense Prediction with Context-Aware Prompting

Unofficial reimplementation of VLA-0 using TRL's SFTTrainer.

An ever-growing playground of notebooks showcasing CLIP's impressive zero-shot capabilities

[IPCAI 2024 (IJCARS special issue)] Surgical-DINO: Adapter Learning of Foundation Models for Depth Estimation in Endoscopic Surgery

Testing adaptation of the DINOv2/3 encoders for vision tasks with Low-Rank Adaptation (LoRA)

Pre-training Reusable Representations for Robotic Manipulation Using Diverse Human Video Data

Open-vocabulary Semantic Segmentation

Orient Anything V2, NeurIPS 2025 Spotlight

Code examples and API documentation for KINOVA® KORTEX™ robotic arms

a minimal, beginner-friendly VLA to show how robot policies can fuse images, text, and states to generate actions

Python bindings for real-time control of Franka Emika robots.

NeurIPS 2025 Spotlight; ICLR2024 Spotlight; CVPR 2024; EMNLP 2024

Run Segment Anything Model 2 on a live video stream

The repository provides code for running inference with the Meta Segment Anything Model 2 (SAM 2) in real-time.

VLM-FO1: Bridging the Gap Between High-Level Reasoning and Fine-Grained Perception in VLMs

EfficientSAM3 compresses SAM3 into lightweight, edge-friendly models via progressive knowledge distillation for fast promptable concept segmentation and tracking.

SAM 3D Objects