five

pick_mask

收藏
Hugging Face2026-06-01 更新2026-06-01 收录
下载链接:
https://huggingface.co/datasets/plzsay/pick_mask
下载链接
链接失效反馈
官方服务:
资源简介:
这是一个用于机器人控制的数据集,专注于机械臂操作任务,可能应用于强化学习或模仿学习。数据集由LeRobot创建,包含30个episodes,总计10770帧数据。数据以parquet格式存储,视频以mp4格式存储。数据集特征包括:动作(action)和观测状态(observation.state),均为6维浮点数组,对应机械臂的六个关节位置(肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置);观测图像(observation.images.cam_left),为左摄像头采集的视频数据,分辨率480x640,帧率30fps,编码格式av1;以及时间戳、帧索引、episode索引、任务索引等元数据。机器人类型为so_follower,数据集总大小约300MB(数据文件100MB,视频文件200MB)。该数据集适用于机器人抓取、操作或其他相关任务的研究和开发。

This is a dataset for robot control, focusing on robotic arm manipulation tasks, potentially applicable to reinforcement learning or imitation learning. The dataset was created by LeRobot, containing 30 episodes with a total of 10770 frames. Data is stored in parquet format, and videos are stored in mp4 format. Dataset features include: action and observation.state, both 6-dimensional float arrays corresponding to the six joint positions of the robotic arm (shoulder translation, shoulder lift, elbow bend, wrist bend, wrist rotation, and gripper position); observation.images.cam_left, video data captured by the left camera with a resolution of 480x640, frame rate of 30fps, and encoding format av1; as well as metadata such as timestamps, frame index, episode index, and task index. The robot type is so_follower, and the total dataset size is approximately 300MB (100MB for data files and 200MB for video files). This dataset is suitable for research and development in robot grasping, manipulation, or other related tasks.
提供机构:
plzsay
创建时间:
2026-06-01
原始信息汇总

数据集概述

  • 数据集名称: pick_mask
  • 许可证: Apache-2.0
  • 任务类别: 机器人学 (robotics)
  • 标签: LeRobot

数据集描述

该数据集使用 LeRobot 创建,可通过 可视化界面 预览。

数据集结构

  • 代码库版本: v3.0
  • 帧率 (fps): 30
  • 机器人类型: so_follower
  • 总集数 (episodes): 30
  • 总帧数 (frames): 10770
  • 总任务数 (tasks): 1
  • 数据文件大小: 100 MB
  • 视频文件大小: 200 MB
  • 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
  • 数据划分: 训练集包含全部30个episode(split: train: "0:30")

特征 (Features)

特征名称 数据类型 形状 描述
action float32 (6,) 6维动作指令(肩部、肘部、腕部、夹爪位置)
observation.state float32 (6,) 6维机器人状态(与action特征名称相同)
observation.images.cam_left video (480, 640, 3) 左侧相机视频流,分辨率480x640,AV1编码,30fps
timestamp float32 (1,) 时间戳
frame_index int64 (1,) 帧索引
episode_index int64 (1,) 集索引
index int64 (1,) 全局索引
task_index int64 (1,) 任务索引

引用

引用信息暂缺([More Information Needed])。

二维码
社区交流群
二维码
科研交流群
商业服务