openarm_bimanual_shirt_folding_high_quality
收藏资源简介:
该数据集是使用LeRobot创建的,专注于机器人任务,具体涉及双臂衬衫折叠的高质量数据。数据集包含100个episodes,共146472帧,帧率为30fps。数据特征包括动作(16维浮点数,代表左右臂各关节和夹爪的位置)、观察状态(16维浮点数,与动作类似)、以及来自多个视角的图像观察:左手腕摄像头(分辨率720x1280,3通道)、右手腕摄像头(分辨率720x1280,3通道)和基础摄像头(分辨率480x640,3通道)。此外,还包括时间戳、帧索引、episode索引、索引和任务索引等元数据。数据集采用Apache 2.0许可证,数据文件格式为parquet,视频文件格式为mp4,机器人类型为双臂开放臂跟随器。数据集用于训练和评估双臂机器人执行衬衫折叠等精细操作任务。
This dataset was developed using LeRobot, focusing on robotic tasks, specifically high-quality data for dual-arm shirt folding tasks. It contains 100 episodes, totaling 146,472 frames at a frame rate of 30 fps. The dataset features include actions (16-dimensional floating-point values representing the positions of each joint and gripper of the left and right arms), observation states (16-dimensional floating-point values, similar to the action space), and image observations from multiple perspectives: left wrist camera (resolution 720×1280, 3 channels), right wrist camera (resolution 720×1280, 3 channels), and base camera (resolution 480×640, 3 channels). Additionally, it includes metadata such as timestamps, frame indices, episode indices, indices and task indices. The dataset is licensed under Apache 2.0, with data files stored in Parquet format and video files in MP4 format. The robotic platform used is the dual-arm open-arm follower. This dataset is designed for training and evaluating dual-arm robots to perform fine manipulation tasks such as shirt folding.
数据集概述
- 数据集名称: openarm_bimanual_shirt_folding_high_quality
- 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 框架创建,专注于双臂机器人折叠衬衫的高质量操作数据。
数据集结构
基本信息
| 项目 | 数值 |
|---|---|
| 总帧数 | 146,472 |
| 总片段数 (episodes) | 100 |
| 总任务数 | 1 |
| 帧率 (FPS) | 30 |
| 机器人类型 | bi_openarm_follower |
| 数据文件大小 | 100 MB |
| 视频文件大小 | 200 MB |
数据特征
-
动作 (action):
- 数据类型: float32
- 形状: (16,)
- 包含: 左臂7个关节位置、左夹爪位置、右臂7个关节位置、右夹爪位置
-
观察状态 (observation.state):
- 数据类型: float32
- 形状: (16,)
- 包含: 与动作相同的16维状态信息(左/右臂关节及夹爪位置)
-
观察图像 (observation.images):
- 左腕相机 (left_wrist): 分辨率 720x1280,3通道,AV1编码,30 FPS
- 右腕相机 (right_wrist): 分辨率 720x1280,3通道,AV1编码,30 FPS
- 基座相机 (base): 分辨率 480x640,3通道,AV1编码,30 FPS
-
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,)
数据分割
- 训练集 (train): 片段索引 0 至 99(共100个片段)
文件结构
- 数据文件路径:
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 - 数据块大小: 1000
引用
目前暂无提供的 BibTeX 引用信息。



