Tactile Gym 2.0: Sim-to-real Deep Reinforcement Learning for Comparing Low-cost High-Resolution Robot Touch
Open-source tactile RL environments with sim-to-real transfer across three low-cost optical sensors (TacTip, DIGIT, DigiTac), for comparing them on the same tasks.开源触觉强化学习环境，支持三种低成本光学触觉传感器（TacTip、DIGIT、DigiTac）的仿真到真机迁移，便于在相同任务上比较。
BibTeX
@misc{lin2022tactile,
title = {{Tactile Gym 2.0: Sim-to-real Deep Reinforcement Learning for Comparing Low-cost High-Resolution Robot Touch}},
author = {Yijiong Lin and John Lloyd and Alex Church and Nathan F. Lepora},
year = {2022},
eprint = {2207.10763},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2207.10763},
}Tactile & Visuotactile Sensing触觉与视触觉感知Sim-to-Real & Simulators仿真与 Sim-to-RealRobot Learning: RL & Imitation机器人学习：强化学习与模仿学习simulatorsim-to-realrltactipdigit