YOLO2431/clean_up_table_v2.1

该数据集是一个机器人技术领域的数据集，使用LeRobot工具创建，专注于双臂机器人（Yam Bimanual）的观测和动作数据。它包含194个episodes，总计148,588帧，覆盖5个不同任务，并附带582个视频。数据以30fps的帧率采集，包括32维的观测状态（如左右机械臂的位置、旋转、夹爪状态和关节角度）和对应的32维动作向量。此外，数据集提供了多视角的视觉信息：头部、左腕和右腕的RGB视频（分辨率720x1280，3通道），以及相应的深度视频（分辨率720x1280，1通道，深度单位为毫米）。所有数据都用于训练，旨在支持机器人学习任务，如模仿学习或强化学习。This dataset is in the field of robotics, created using the LeRobot tool, focusing on observation and action data for a bimanual robot (Yam Bimanual). It contains 194 episodes, totaling 148,588 frames, covering 5 distinct tasks, and includes 582 videos. The data is collected at 30 fps and includes 32-dimensional observation states (such as left and right arm positions, rotations, gripper status, and joint angles) and corresponding 32-dimensional action vectors. Additionally, the dataset provides multi-view visual information: RGB videos from the head, left wrist, and right wrist (resolution 720x1280, 3 channels), as well as corresponding depth videos (resolution 720x1280, 1 channel, depth unit in millimeters). All data is used for training, aiming to support robotics learning tasks such as imitation learning or reinforcement learning.