Skip to content

Dataset · Renton-Ren

Robot learning images from the XIRL paper (CoRL 2021)

Renton-Ren/multi-task-tcc-robosuite

Visual data from the Robosuite simulator for reproducing reinforcement learning robot experiments (XIRL).

What's inside

The dataset contains between 10,000 and 100,000 images accompanying the code publication for the XIRL (Cross-embodiment Inverse Reinforcement Learning, CoRL 2021) paper by researchers from Robotics at Google and Stanford. The materials are designed to reproduce skill transfer experiments between different robots in the Robosuite environment.

What you can do with it

  • Training inverse reinforcement learning (Inverse RL) models for robotics
  • Reproducing XIRL paper results and testing skill transfer between different manipulators
Size
10K–100K rows
Downloads this month
28.8K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("Renton-Ren/multi-task-tcc-robosuite")