Recent Talk

Talks by Joseph (PI)

Invited Talk @ RSS 2023

Interdisciplinary Exploration of Generalizable Manipulation Policy Learning Workshop

Talks by Students

Jesse Zhang (CoRL 2023 Paper Session)

Minho Heo (RSS 2023 Paper Session)

Environments / Datasets

Robotics

Open X-Embodiment Dataset
The Open X-Embodiment Dataset, the most extensive open-source dataset for real robots available, offers over 1M+ million real robot trajectories across 22 different robot embodiments.
[Project] [Dataset]

FurnitureBench / FurnitureSim (Best System Paper Award at RSS 2023)
FurnitureBench is a reproducible real-world furniture assembly benchmark for long-horizon complex robotic manipulation. Its simulated environment, FurnitureSim, is a fast and realistic simulator based on Isaac Gym and Factory.
[PDF] [Project] [Code] [Dataset]

CLVR Jaco Play Dataset
The CLVR Jaco Play Dataset provides 1,085 teleoperated robotic episodes with accompanying language annotations.
[Dataset]

IKEA Furniture Assembly Environment
The IKEA Furniture Assembly Environment is one of the first benchmarks for testing and accelerating the automation of complex manipulation tasks.
[PDF] [Project] [Code]

Generalization

CREATE: Chain REAction Tool Environment
CREATE is a reinforcement learning benchmark for solving a class of complex physics puzzles with diverse tools.
[PDF] [Project] [Code] [Demo]