Bimanual Tasks

该数据集包含四个不同的双手操作任务：胡椒磨、夹克拉链、记号笔盖和切蔬菜，旨在评估BUDS模型在动态重新稳定场景中的性能。该数据集使得研究者能够评估不同双手任务的成功率，强调了对高精度动作策略和动态稳定策略的需求。这一评估涵盖了四个任务，并具有对未见对象的泛化能力，这些任务均需要双手协调与稳定性的操作。

This dataset includes four distinct bimanual manipulation tasks: pepper grinding, zippering a jacket, removing a marker cap, and vegetable cutting. It is designed to evaluate the performance of the BUDS model in dynamic restabilization scenarios. The dataset enables researchers to assess the success rates of various bimanual tasks, emphasizing the need for high-precision motion policies and dynamic stabilization policies. This evaluation covers these four tasks and features generalization capability to unseen objects, all of which require coordinated and stable bimanual manipulation.