SoftGym
收藏资源简介:
SoftGym是由卡内基梅隆大学创建的一套开源模拟基准数据集,专注于可变形物体操作,包含10个涉及绳索、布料和可变属性流体的挑战性环境。数据集通过标准的OpenAI Gym API和Python接口创建新环境,旨在通过提供标准化的环境集,促进可变形物体操作领域的新算法开发和比较。SoftGym的应用领域广泛,包括日常生活中的衣物折叠、食品制作以及工业应用中的包装或处理电缆等,旨在解决机器人操作可变形物体的高维状态表示和复杂动力学问题。
SoftGym is an open-source simulated benchmark dataset developed by Carnegie Mellon University, focusing on deformable object manipulation. It includes 10 challenging environments involving ropes, fabrics, and fluids with variable physical properties. The benchmark supports creating new environments via the standard OpenAI Gym API and Python interface, aiming to facilitate the development and comparison of novel algorithms in the field of deformable object manipulation by providing a standardized set of environments. SoftGym has broad application scenarios, including daily tasks such as cloth folding and food preparation, as well as industrial applications like packaging or cable handling. It aims to address the challenges of high-dimensional state representation and complex dynamics in robotic manipulation of deformable objects.




