[CoRL '23] Dexterous piano playing with deep reinforcement learning.

Code for PerAct², a language-conditioned imitation learning agent designed for bimanual robotic manipulation using the RLBench environment. It includes dataset generation, training scripts, and evaluation tools for 6-DoF control of two robot arms across complex tasks.

The New Dexterity Open-Bimanual-Manipulator project

Code for paper: "Look, Focus, Act: Efficient and Robust Robot Learning via Human Gaze and Foveated Vision Transformers"

Official implementation of LLM+MAP: Bimanual Robot Task Planning using Large Language Models (LLMs) and Planning Domain Definition Language (PDDL). Codes and files are coming soon.

This repository is for tracking progress of bimanual manipulation for different tasks.

MANIP-ULTRADEX — Bimanual Dexterous Grasping

RoboPianist with tactile feedback and velocity-based evaluation for piano playing.

Identifying actions belonging to the Uncontrolled Manifold for bimanual manipulation of a cursor using reinforcement learning.

A tested control, planning, perception, bimanual, and learning platform for the Enactic OpenArm v2 (7-DOF x2) in MuJoCo: IK / Cartesian / compliant control, RRT-Connect planning, dynamic catching and throwing, articulated / cloth manipulation, language-commanded skills, a learned ACT vision policy, RL insertion, and OpenArm-Bench.

Geodude bimanual robot control library