eval_test_10
收藏资源简介：
该数据集是一个用于机器人学任务的数据集，使用LeRobot工具创建。数据集包含机器人动作、状态观测（包括关节位置如肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置）、顶部和前置摄像头视频（分辨率为480x640，帧率为30fps，视频编码为av1），以及时间戳、帧索引、episode索引等元数据。数据以parquet格式存储，总共有1个episode、914帧、1个任务，适用于机器人控制和模仿学习等应用。
This dataset is designed for robotics tasks and was created using the LeRobot tool. It includes robot actions, state observations (such as joint positions like shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position), top and front camera videos (with a resolution of 480x640, a frame rate of 30fps, and video encoding as av1), as well as metadata such as timestamps, frame indices, and episode indices. The data is stored in parquet format, with a total of 1 episode, 914 frames, and 1 task, making it suitable for applications like robot control and imitation learning.
数据集概述
- 数据集名称: eval_test_10
- 许可证: Apache-2.0
- 任务类别: 机器人 (Robotics)
- 标签: LeRobot
数据集结构
- 代码库版本: v3.0
- 机器人类型: so_follower
- 总片段数 (episodes): 1
- 总帧数 (frames): 914
- 总任务数 (tasks): 1
- 块大小 (chunks_size): 1000
- 数据文件大小: 100 MB
- 视频文件大小: 200 MB
- 帧率 (fps): 30
- 数据分割: 仅包含训练集 (train)，范围 "0:1"
数据特性 (Features)
| 特性名称 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
| action | float32 | [6] | 机器人动作，包含6个关节位置 (shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper) |
| observation.state | float32 | [6] | 机器人观测状态，同样包含6个关节位置 |
| observation.images.top | video | (480, 640, 3) | 顶部摄像头视频，编码为AV1，分辨率480x640，30fps |
| observation.images.front | video | (480, 640, 3) | 前置摄像头视频，编码为AV1，分辨率480x640，30fps |
| timestamp | float32 | [1] | 时间戳 |
| frame_index | int64 | [1] | 帧索引 |
| episode_index | int64 | [1] | 片段索引 |
| index | int64 | [1] | 通用索引 |
| task_index | int64 | [1] | 任务索引 |
数据存储路径
- 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
其他信息
- 该数据集使用 LeRobot 框架创建。
- 可视化链接: Visualize this dataset
