rollout_diffusion_mir1_ddim_nodisplay_20260706_124028
收藏Hugging Face2026-07-06 更新2026-07-06 收录
下载链接:
https://huggingface.co/datasets/Akshit03/rollout_diffusion_mir1_ddim_nodisplay_20260706_124028
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是使用LeRobot创建的机器人学数据集,专门用于机器人控制与视觉任务。它包含3个总片段,共4691帧数据,数据文件大小为100MB,视频文件大小为200MB。数据集特征包括动作(6个关节位置: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)、观察状态(与动作相同的6个关节位置)、两个摄像头图像(顶部和腕部摄像头,均为RGB视频,分辨率480x640,帧率30fps),以及时间戳、帧索引、片段索引、任务索引等元数据。机器人类型为so_follower,适用于训练和评估机器人模仿学习或强化学习模型。数据集以Parquet格式存储,视频以MP4格式编码为AV1,并包含训练分割。
This dataset is a robotics dataset created using LeRobot, specifically designed for robot control and vision tasks. It contains 3 total segments with 4691 frames, a data file size of 100MB, and a video file size of 200MB. The dataset features include actions (6 joint positions: shoulder translation, shoulder elevation, elbow flexion, wrist flexion, wrist rotation, and gripper position), observation states (the same 6 joint positions as actions), two camera images (top and wrist cameras, both RGB videos with a resolution of 480x640 and a frame rate of 30fps), as well as metadata such as timestamps, frame index, segment index, and task index. The robot type is so_follower, suitable for training and evaluating robot imitation learning or reinforcement learning models. The dataset is stored in Parquet format, with videos encoded in MP4 as AV1, and includes a training split.
提供机构:
Akshit03创建时间:
2026-07-06
原始信息汇总
数据集概述
- 数据集名称:
rollout_diffusion_mir1_ddim_nodisplay_20260706_124028 - 许可证: Apache-2.0
- 任务类别: 机器人学 (robotics)
- 创建工具: 基于 LeRobot 框架创建
数据集结构
基本信息
| 属性 | 值 |
|---|---|
| codebase_version | v3.0 |
| fps (帧率) | 30 |
| 总片段数 (total_episodes) | 3 |
| 总帧数 (total_frames) | 4691 |
| 总任务数 (total_tasks) | 1 |
| 数据文件大小 | 100 MB |
| 视频文件大小 | 200 MB |
| 机器人类型 | so_follower |
| 数据分割 (splits) | 仅训练集 (train: 0:3) |
数据特征
- action (动作): 6维
float32向量,包含机器人关节位置(shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper) - observation.state (观测状态): 6维
float32向量,结构与 action 相同 - observation.images.top (顶部相机图像): 视频数据,分辨率 480x640,3通道,编码格式为 AV1,帧率 30 fps
- observation.images.wrist (腕部相机图像): 视频数据,分辨率 480x640,3通道,编码格式为 AV1,帧率 30 fps
- timestamp: 时间戳,
float32,单值 - frame_index: 帧索引,
int64,单值 - episode_index: 片段索引,
int64,单值 - index: 索引,
int64,单值 - task_index: 任务索引,
int64,单值
文件存储路径
- 数据文件 (parquet):
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet(分块大小: 1000) - 视频文件 (mp4):
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4
引用说明
- 引用格式: 目前尚未提供 BibTeX 引用信息([More Information Needed])




