gsethia08/part2-insert_remove_furniture_bench_round_table

该数据集是一个用于机器人学习的数据集，由LeRobot工具创建。数据集专注于手部机器人操作任务，包含4个不同任务，总共有4892个episodes和696178帧数据，帧率为30fps。数据格式包括parquet数据文件和mp4视频文件。特征涵盖观测数据（如1080x1920分辨率的RGB相机图像、48维机器人状态向量）、相机内外参数、48维动作向量、语言指令、时间戳、帧索引、episode索引等。数据集适用于机器人控制、模仿学习或强化学习任务，旨在支持手部机器人的多任务操作学习。This dataset is a robotics learning dataset created using the LeRobot tool. It focuses on hand robot manipulation tasks, containing 4 distinct tasks with a total of 4892 episodes and 696178 frames at 30fps. The data format includes parquet data files and mp4 video files. Features encompass observations (such as RGB camera images at 1080x1920 resolution, 48-dimensional robot state vectors), camera intrinsic and extrinsic parameters, 48-dimensional action vectors, language instructions, timestamps, frame indices, episode indices, etc. The dataset is suitable for robot control, imitation learning, or reinforcement learning tasks, aiming to support multi-task manipulation learning for hand robots.