UR5e_sim_pick_place_can_1000_absolute_EEF_actions
收藏资源简介:
这是一个用于机器人拾放任务的模拟数据集,基于LeRobot框架在robosuite环境中创建。数据集针对UR5e机器人,通过脚本化策略执行拾放操作,具体对象为can。数据集包含1000个episodes和130165帧数据,以20fps的帧率采集。数据特征包括:机器人的动作(4维浮点数组)、状态观测(13维浮点数组)、来自两个摄像头的图像观测(每个为256x256x3的视频数据,使用AV1编码),以及时间戳、帧索引、episode索引等元数据。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200MB。该数据集适用于机器人学习任务,如强化学习或模仿学习,专注于绝对末端执行器(EEF)动作控制。
This is a simulated dataset for robotic pick-and-place tasks, developed in the Robosuite environment based on the LeRobot framework. Targeting the UR5e robot, the dataset performs pick-and-place operations using scripted policies, with the target object being a can. The dataset contains 1000 episodes and 130,165 frames of data, collected at a frame rate of 20 FPS. The data features include: robot actions (4-dimensional floating-point arrays), state observations (13-dimensional floating-point arrays), image observations from two cameras (each being 256×256×3 video data encoded with AV1), as well as metadata such as timestamps, frame indices, and episode indices. The data is stored in Parquet file format, with a total size of 100 MB for the data files and 200 MB for the video files. This dataset is suitable for robotic learning tasks such as reinforcement learning or imitation learning, with a focus on absolute end-effector (EEF) action control.
数据集概述
- 数据集名称:UR5e_sim_pick_place_can_1000_absolute_EEF_actions
- 数据集地址:https://huggingface.co/datasets/castanetnicolas/UR5e_sim_pick_place_can_1000_absolute_EEF_actions
- 许可证:Apache-2.0
- 任务类别:机器人学(robotics)
- 标签:LeRobot, ur5e, robosuite, pick-place, scripted
数据集描述
该数据集基于 LeRobot 框架创建,包含使用 UR5e 机器人模拟执行拾取和放置罐头的任务,共 1000 个预设脚本化轨迹。
数据集结构
- 数据格式:包含 Parquet 文件和视频文件。
- 帧率 (fps):20
- 总片段数 (episodes):1000
- 总帧数 (frames):130165
- 总任务数 (tasks):1
- 数据文件大小:约 100 MB
- 视频文件大小:约 200 MB
- 机器人类型:ur5e
特征列表
| 特征名称 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
| action | float32 | (4,) | 动作,包含 4 维浮点数 |
| observation.state | float32 | (13,) | 观测状态,包含 13 维浮点数 |
| observation.images.camera1 | video | (256, 256, 3) | 摄像头1图像,分辨率256x256,3通道,AV1编码 |
| observation.images.camera2 | video | (256, 256, 3) | 摄像头2图像,分辨率256x256,3通道,AV1编码 |
| timestamp | float32 | (1,) | 时间戳 |
| frame_index | int64 | (1,) | 帧索引 |
| episode_index | int64 | (1,) | 片段索引 |
| index | int64 | (1,) | 索引 |
| task_index | int64 | (1,) | 任务索引 |
数据路径
- 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
数据集划分
- 训练集 (train):索引 0 至 999(共 1000 个片段)
引用信息
该数据集的 BibTeX 引用信息待补充。



