libero
收藏资源简介:
该数据集是一个机器人学数据集,使用LeRobot创建。数据集包含1693个总剧集,273465个总帧,覆盖40个任务。数据以parquet文件格式存储,总数据文件大小为100MB,视频文件大小为500MB。帧率为10fps。数据集特征包括两个视频观察(image和image2,均为256x256x3分辨率)、状态观察(8维浮点数组)、动作(7维浮点数组),以及时间戳、帧索引、剧集索引、索引和任务索引等元数据。数据集使用Apache 2.0许可证,适用于机器人任务训练和研究。
This dataset is a robotics dataset created using LeRobot. It contains a total of 1693 episodes and 273465 frames, covering 40 tasks. The data is stored in parquet file format, with a total data file size of 100MB and a video file size of 500MB. The frame rate is 10fps. Dataset features include two video observations (image and image2, both with a resolution of 256x256x3), state observations (an 8-dimensional floating-point array), actions (a 7-dimensional floating-point array), as well as metadata such as timestamps, frame indices, episode indices, indices, and task indices. The dataset uses the Apache 2.0 license and is suitable for robot task training and research.
数据集概述
- 名称: libero
- 许可证: Apache-2.0
- 任务类别: 机器人学(robotics)
- 标签: LeRobot
数据集规模
- 总片段数(episodes): 1693
- 总帧数(frames): 273,465
- 总任务数(tasks): 40
- 帧率(fps): 10.0
- 数据文件大小: 100 MB
- 视频文件大小: 500 MB
数据集结构
- 机器人类型: panda
- 数据格式: Parquet 文件(路径:
data/*/*.parquet) - 视频格式: MP4 文件(路径:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4) - 分割: 全部数据用于训练(train: 0:1693)
特征
| 特征名 | 数据类型 | 形状 | 描述 |
|---|---|---|---|
| observation.images.image | 视频 | [256, 256, 3] | 主摄像头图像(高x宽x通道),帧率10fps,AV1编码,yuv420p像素格式 |
| observation.images.image2 | 视频 | [256, 256, 3] | 第二个摄像头图像(同主摄像头规格) |
| observation.state | float32 | [8] | 机器人状态(8维向量) |
| action | float32 | [7] | 机器人动作(7维向量) |
| 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 |
创建工具
该数据集使用 LeRobot 创建。



