Three_Cubes_2
收藏资源简介:
该数据集是一个专为机器人控制任务设计的数据集,由LeRobot开发。它包含100个剧集,总计50790帧,帧率为30fps。数据集提供多模态观测数据,包括来自前、右和手腕摄像头的RGB视频,分辨率为480x640,以及机器人的状态信息(如关节位置)。动作数据涵盖6个自由度,具体包括肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部滚动和夹爪位置。数据以Parquet文件格式存储,总数据文件大小为100MB,视频文件大小为200MB。该数据集适用于机器人学习和强化学习研究,支持模拟或真实机器人的控制任务。
This dataset is designed for robot control tasks, created by LeRobot. It contains 100 episodes, totaling 50790 frames at a frame rate of 30fps. The dataset provides multimodal observations, including RGB videos from front, right, and wrist cameras with a resolution of 480x640, along with robot state information (such as joint positions). Action data includes 6 degrees of freedom, specifically shoulder translation, shoulder lift, elbow bend, wrist bend, wrist roll, and gripper position. The data is stored in Parquet file format, with a total data file size of 100MB and video file size of 200MB. It is suitable for robot learning and reinforcement learning research, supporting control tasks for simulated or real robots.
数据集概述
- 数据集名称: Three_Cubes_2
- 许可证: Apache-2.0
- 任务类别: 机器人学 (robotics)
- 数据集创建工具: LeRobot
数据集规模
- 总片段数 (Episodes): 100
- 总帧数 (Frames): 50,790
- 总任务数: 1
- 帧率 (FPS): 30
- 数据文件大小: 100 MB
- 视频文件大小: 200 MB
数据划分
- 训练集 (Train): 片段索引 0 到 99(共 100 个片段)
机器人类型
- 机器人型号: so_follower
特征结构
动作 (Action)
- 数据类型: float32
- 维度: 6
- 动作名称: shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos, wrist_flex.pos, wrist_roll.pos, gripper.pos
观察状态 (Observation State)
- 数据类型: float32
- 维度: 6
- 状态名称: shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos, wrist_flex.pos, wrist_roll.pos, gripper.pos
图像观察 (Observation Images)
- 摄像头视角: 前方 (front)、右侧 (right)、腕部 (wrist)
- 图像尺寸: 480x640 像素,3 通道
- 视频编码: AV1
- 视频帧率: 30 FPS
- 色彩格式: yuv420p
- 非深度图: 是
其他特征
- 时间戳 (timestamp): float32,形状 [1]
- 帧索引 (frame_index): int64,形状 [1]
- 片段索引 (episode_index): int64,形状 [1]
- 索引 (index): int64,形状 [1]
- 任务索引 (task_index): int64,形状 [1]
数据存储
- 数据路径:
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet - 视频路径:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4 - 块大小 (chunks_size): 1000



