fecasado/burger-to-plate-320x240

该数据集是使用LeRobot工具创建的机器人学习数据集，适用于机器人控制任务。数据集包含90个episodes和134,493帧数据，以15fps的帧率采集。数据特征包括：动作向量（26维，表示左右手臂的线性/角速度、手部关节控制等）、状态观测（55维，包括左右手臂和手部关节的位置和力矩，以及视线坐标）、多视角图像视频（左、右、用户和用户注视视角，每个视频为240x320分辨率、3通道、AV1编码）。此外，还包含时间戳、帧索引、episode索引等元数据。数据集结构基于parquet文件和视频文件，总数据大小约300MB（数据100MB，视频200MB），机器人类型为blueberry_ros。

This dataset is a robotics learning dataset created using the LeRobot tool, suitable for robot control tasks. It contains 90 episodes and 134,493 frames, collected at 15fps. Features include: action vectors (26-dimensional, representing linear/angular velocities of left/right arms, hand joint controls, etc.), state observations (55-dimensional, including positions and efforts of left/right arm and hand joints, and gaze coordinates), multi-view image videos (left, right, user, and user gaze views, each with 240x320 resolution, 3 channels, AV1 encoding). Additionally, metadata such as timestamps, frame indices, and episode indices are included. The dataset structure is based on parquet and video files, with a total size of approximately 300MB (100MB for data, 200MB for videos), and the robot type is blueberry_ros.