DECO: Decoupled Multimodal Diffusion Transformer for Bimanual Dexterous Manipulation with a Plugin Tactile Adapter DECO 50 is a bimanual dexterous manipulation dataset with tactile sensing, comprising 50 hours of teleoperated data across 4 scenarios and 28 subtasks, totaling over 5 million frames collected on real dual arm robots. Dataset Structure Episode Label Format (data.pkl) Field Description index Index of the data sample left action Left arm and hand action sequence right action Right arm and hand action sequence head action Active head action sequence left obs Left arm and hand observation sequence right obs Right arm and hand observation sequence head obs Active head observation sequence condition indx Index of the sub task for one hot embedding Using the Dataset You can download the dataset and use it directly with our training scripts. Please note the following: Task1–Task4 correspond to four manipulation scenarios: Pick and Place, Material Sorting, Waste Disposal, and Assembly. Each scenario is further divided into multiple subtasks. Within each subtask, the file data.pkl uses condition indx to distinguish different object instances and task types. In our paper and rele…
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy