CALVIN: A Benchmark for Language-Conditioned Policy Learning for Long-Horizon Robot Manipulation

摘要

An open-source simulated benchmark for learning long-horizon language-conditioned robotic manipulation tasks.