lelab-test_20260511_125014
收藏Hugging Face2026-05-11 更新2026-05-11 收录
下载链接:
https://huggingface.co/datasets/nikodembartnik/lelab-test_20260511_125014
下载链接
链接失效反馈官方服务:
资源简介:
该数据集使用LeRobot创建,是一个机器人控制数据集,专门用于机器人学习任务。数据集包含5个完整的情节(episodes),总计1657帧数据,采样频率为30 FPS。数据特征包括:6维浮点型动作向量(控制机械臂的肩部、肘部、腕部和夹持器位置)、6维浮点型状态观测向量(与动作向量结构相同)、顶部摄像头图像(480x640分辨率,RGB三通道,视频编码为AV1,无音频),以及时间戳、帧索引、情节索引、索引和任务索引等元数据。数据集总大小约为300 MB(数据文件100 MB,视频文件200 MB),使用Apache 2.0许可证。机器人类型为so_follower,数据以Parquet格式存储,视频以MP4格式存储,并包含训练分割(所有5个情节用于训练)。
This dataset is a robotic control dataset developed for robot learning tasks, constructed using the LeRobot framework. It contains 5 full episodes, totaling 1657 frames with a sampling rate of 30 FPS. The dataset's data features include: a 6-dimensional floating-point action vector used to control the positions of the robotic arm's shoulder, elbow, wrist, and gripper, a 6-dimensional floating-point state observation vector sharing the same structure as the action vector, top-down camera images with 480×640 resolution, RGB three-channel, encoded in AV1 format with no audio, as well as metadata such as timestamps, frame index, episode index, global index, and task index. The total size of the dataset is approximately 300 MB, comprising 100 MB of data files and 200 MB of video files, and it is released under the Apache 2.0 open-source license. The target robotic platform is so_follower. The structured data is stored in Parquet format while the video files are stored in MP4 format. The dataset includes a training split, where all 5 episodes are used for model training.
提供机构:
nikodembartnik
创建时间:
2026-05-11
原始信息汇总
数据集概述
该数据集是一个用于机器人操作的录制数据集,基于LeRobot框架创建。
- 数据集名称:
nikodembartnik/lelab-test_20260511_125014 - 任务类型: 机器人学 (robotics)
- 许可证: Apache-2.0
- 总帧数: 1657帧
- 总片段数: 5个片段
- 总任务数: 1个任务
- 帧率: 30 FPS
- 机器人类型:
so_follower - 数据文件大小: 100 MB(Parquet文件)
- 视频文件大小: 200 MB(MP4文件)
数据特征
数据集包含以下特征:
| 特征名称 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
action |
float32 | (6,) | 6维动作指令,包括机械臂各关节位置和夹爪位置 |
observation.state |
float32 | (6,) | 6维观测状态,与动作空间一致 |
observation.images.top |
video | (480, 640, 3) | 顶部摄像头录制的高清视频(AV1编码,30 FPS) |
timestamp |
float32 | (1,) | 时间戳 |
frame_index |
int64 | (1,) | 帧索引 |
episode_index |
int64 | (1,) | 片段索引 |
index |
int64 | (1,) | 全局索引 |
task_index |
int64 | (1,) | 任务索引 |
动作/状态维度含义
shoulder_pan.posshoulder_lift.poselbow_flex.poswrist_flex.poswrist_roll.posgripper.pos
数据集划分
- 训练集: 全部5个片段(索引0:4)
视频规格
- 分辨率: 480×640像素
- 编码格式: AV1
- 像素格式: yuv420p
- 通道数: 3(RGB)
- 帧率: 30 FPS
数据存储结构
- Parquet数据路径:
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet - 视频文件路径:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4 - 元数据文件:
meta/info.json



