DPD: A Dataset Simultaneous Capture Motion and Tactile Information Using Wearables
This dataset includes human pose data from an IMU sensor network, hand pose data from a smart data glove, and palm contact force data from a pressure sensing glove.
Estimating human body posture is critical for enabling human-robot collaboration, preventing occupational diseases, and training humanoid robots. Recent advancements in wearable sensors, such as Inertial Measurement Units (IMUs), have made it possible to conduct motion capture research in outdoor and factory settings. However, these sensors do not capture force information, which is a vital parameter for object manipulation. To address this issue, we propose the DPD (Dynamics Pose Dataset) multimodal dataset that provides missing force information during hand-object/tool manipulation. This dataset includes human pose data from an IMU sensor network, hand pose data from a smart data glove, and palm contact force data from a pressure sensing glove. In this paper, we present the hardware setup, experimental protocol, data format, and detailed data. We believe that this dataset can provide urgently needed information for medical and robotics research.
