robosuite
收藏资源简介:
robosuite是一个基于MuJoCo物理引擎的机器人学习模拟框架,旨在通过模块化设计支持机器人任务的创建和一系列基准环境的提供,以促进数据驱动型机器人算法的研究。该框架源自SURREAL项目,现已成为Advancing Robot Intelligence through Simulated Environments (ARISE) Initiative的一部分。robosuite v1.0包含七种机器人模型、八种夹持器模型、六种控制器模式和九种标准化任务,支持多种传感器和人类演示数据的收集,适用于解决机器人控制和操作中的复杂问题。
robosuite is a robotics learning simulation framework built on the MuJoCo physics engine. It is designed to support the creation of robotic tasks and provide a suite of benchmark environments via modular design, thereby advancing research in data-driven robotic algorithms. Originating from the SURREAL project, this framework has now become part of the Advancing Robot Intelligence through Simulated Environments (ARISE) Initiative. robosuite v1.0 includes seven robot models, eight gripper models, six controller modes, and nine standardized tasks. It supports the collection of data from various sensors and human demonstrations, and is applicable to solving complex problems in robotic control and manipulation.




