autoresearch-vibepi-000255-cucumber-grab
收藏资源简介：
该数据集由LeRobot创建，专注于机器人操作任务。它包含机器人状态观测（如肩部、肘部、腕部等关节的位置）、来自三个不同视角的图像观测（基础、顶部和腕部摄像头，均为RGB视频，分辨率480x640，帧率20fps），以及对应的动作数据。此外，还包括时间戳、帧索引、剧集索引、任务索引等元数据。数据集总共有1个剧集、166帧和1个任务，数据以parquet文件格式存储，视频以mp4格式存储，适用于机器人学习和控制研究。机器人类型为vibeboard_follower_tilt。
This dataset is created by LeRobot and focuses on robot manipulation tasks. It includes robot state observations (such as positions of joints like shoulder, elbow, and wrist), image observations from three different perspectives (base, top, and wrist cameras, all in RGB video format with a resolution of 480x640 and a frame rate of 20fps), and corresponding action data. Additionally, it contains metadata such as timestamps, frame indices, episode indices, and task indices. The dataset consists of 1 episode, 166 frames, and 1 task, with data stored in parquet file format and videos in mp4 format, suitable for robot learning and control research. The robot type is vibeboard_follower_tilt.
数据集概述
- 数据集名称:
autoresearch-vibepi-000255-cucumber-grab - 创建工具: LeRobot
- 许可证: Apache-2.0
- 任务类别: 机器人技术 (robotics)
- 标签: LeRobot
数据集结构
基本信息
- 代码库版本: v3.0
- 帧率 (FPS): 20
- 机器人类型:
vibeboard_follower_tilt
数据规模
- 总片段数 (total_episodes): 1
- 总帧数 (total_frames): 166
- 总任务数 (total_tasks): 1
- 数据文件大小: 100 MB (parquet)
- 视频文件大小: 200 MB (mp4)
- 数据分块大小: 1000
特征 (Features)
| 特征名称 | 数据类型 | 形状 | 描述 |
|---|---|---|---|
observation.state |
float32 | (7,) | 机器人状态: shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper, tilt 的位置 |
observation.images.base |
video | (480, 640, 3) | 基础摄像头图像, 编码格式 av1, 20 FPS |
observation.images.top |
video | (480, 640, 3) | 顶部摄像头图像, 编码格式 av1, 20 FPS |
observation.images.wrist |
video | (480, 640, 3) | 腕部摄像头图像, 编码格式 av1, 20 FPS |
action |
float32 | (7,) | 动作: shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper, tilt 的位置 |
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 |
数据分割
- 训练集 (train):
0:1(所有数据用于训练)
文件路径格式
- 数据文件:
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet - 视频文件:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4
