Colosseum

A Benchmark for Evaluating Generalization for Robotic Manipulation

Wilbert Pumacay^*, Ishika Singh^*, Jiafei Duan^*, Ranjay Krishna, Jesse Thomason, Dieter Fox

This repository contains the implementation of the benchmark mentioned in the title. The code for the benchmark builds on top of PyRep, and integrates with RLBench, implementing 20 out of the original 100 tasks.

Documentation can be found here

🚧 REPOSITORY UNDER DEVELOPMENT 🚧
We're migrating our codebase to this repository and adding documentation for the usage of the benchmark. We'll make a public release once we're done.

Citation

@article{pumacay2024colosseum,
  title     = {THE COLOSSEUM: A Benchmark for Evaluating Generalization for Robotic Manipulation},
  author    = {Pumacay, Wilbert and Singh, Ishika and Duan, Jiafei and Krishna, Ranjay and Thomason, Jesse and Fox, Dieter},
  booktitle = {arXiv preprint arXiv:2402.08191},
  year      = {2024},
}