遇见数据集

so101_socket_pick_and_place

收藏
Hugging Face2026-07-06 更新2026-07-06 收录
官方服务:

资源简介:

该数据集是一个机器人学数据集,专门用于机器人跟随任务。它包含3个完整的episode,总计1614帧数据,涉及1个特定任务。数据集由LeRobot创建,采用Apache 2.0许可证。数据以parquet文件格式存储,视频文件为MP4格式。数据集的特征包括:机器人的动作空间(6个关节位置:肩部平移、肩部升降、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置);状态观测(同样为6个关节位置);图像观测包括前视摄像头和腕部摄像头的视频流,分辨率均为480x640,彩色三通道,帧率为30fps;此外,还包含时间戳、帧索引、episode索引、任务索引等元数据。机器人类型为so_follower,表明数据可能用于机器人跟随或模仿学习任务。数据集总数据文件大小约为100MB,视频文件大小约为200MB。

This is a robotics dataset dedicated to robot following tasks. It contains 3 complete episodes, totaling 1614 frames, and involves 1 specific task. The dataset was created by LeRobot and is licensed under Apache 2.0. The data is stored in Parquet file format, with video files in MP4 format. The dataset's features include: the robot's action space (6 joint positions: shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, and gripper position); state observations (also consisting of 6 joint positions); image observations include video streams from the front-facing camera and wrist camera, both with a resolution of 480×640, 3-channel color, and a frame rate of 30fps; additionally, it includes metadata such as timestamps, frame indices, episode indices, and task indices. The robot type is so_follower, which indicates that the dataset may be used for robot following or imitation learning tasks. The total size of the dataset's data files is approximately 100 MB, while the video files are approximately 200 MB.

提供机构:
chenproject35
创建时间:
2026-07-06
原始信息汇总

数据集总览

  • 名称: so101_socket_pick_and_place
  • 许可协议: Apache-2.0
  • 任务类别: 机器人(Robotics)
  • 标签: LeRobot
  • 创建工具: 基于 LeRobot 构建

数据集规模

  • 总片段数: 3
  • 总帧数: 1614
  • 总任务数: 1
  • 数据文件大小: 100 MB
  • 视频文件大小: 200 MB
  • 帧率: 30 FPS

数据集划分

  • 训练集: 片段 0 到 2(共3个片段)

机器人类型

  • 机器人型号: so_follower

数据特征

数据存储为 Parquet 和 MP4 格式,文件路径规则:

  • 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

动作(action)

  • 数据类型: float32
  • 维度: 6
  • 变量名: shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos, wrist_flex.pos, wrist_roll.pos, gripper.pos

观测状态(observation.state)

  • 数据类型: float32
  • 维度: 6
  • 变量名: shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos, wrist_flex.pos, wrist_roll.pos, gripper.pos

观测图像(observation.images)

  • 前摄像头 (front):
    • 分辨率: 480×640
    • 通道数: 3
    • 编码: AV1
    • 像素格式: yuv420p
    • 帧率: 30 FPS
    • 非深度图: 是
  • 腕部摄像头 (wrist):
    • 分辨率: 480×640
    • 通道数: 3
    • 编码: AV1
    • 像素格式: yuv420p
    • 帧率: 30 FPS
    • 非深度图: 是

其他特征

  • timestamp: float32,形状 [1]
  • frame_index: int64,形状 [1]
  • episode_index: int64,形状 [1]
  • index: int64,形状 [1]
  • task_index: int64,形状 [1]
二维码
社区交流群
二维码
科研交流群
商业服务