Three_Cubes_1
收藏资源简介:
该数据集是一个机器人控制数据集,使用LeRobot工具创建,专为机器人学习任务设计。数据集包含100个episodes,总计51387帧,涉及单一任务。数据以.parquet格式存储,总大小为100MB,同时包含视频文件,总大小为200MB,帧率为30fps。数据集的特征包括动作(action)和观察(observation),其中动作由6个浮点值表示,对应机器人的关节位置(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)。观察部分包括状态(state)和图像(images),状态同样由6个浮点值表示关节位置;图像数据来自三个视角(前视、右侧、腕部),每个视角为480x640分辨率、3通道的RGB视频,使用av1编码和yuv420p像素格式。此外,数据集还包含时间戳、帧索引、episode索引、索引和任务索引等元数据。数据集适用于机器人控制、强化学习或计算机视觉任务的研究和开发。
This is a robotics control dataset created using the LeRobot tool, specifically designed for robot learning tasks. It contains 100 episodes totaling 51387 frames for a single task. The data is stored in .parquet format with a total size of 100 MB, and also includes video files with a total size of 200 MB at a frame rate of 30 fps. The dataset features two core components: action and observation. The action is represented by 6 floating-point values, corresponding to the robot's joint positions including shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, and gripper position. The observation includes state and images. The state is also represented by 6 floating-point values for joint positions. The image data originates from three perspectives: front view, right side view, and wrist view. Each perspective provides 480x640 resolution, 3-channel RGB videos, encoded with AV1 and using the yuv420p pixel format. Additionally, the dataset includes metadata such as timestamps, frame indices, episode indices, indices, and task indices. This dataset is suitable for research and development in robotics control, reinforcement learning, or computer vision tasks.
数据集概述:Three_Cubes_1
- 许可协议:Apache 2.0
- 任务类别:机器人学 (Robotics)
- 标签:LeRobot
- 创建工具:基于 LeRobot 构建
数据集结构
- 代码版本:v3.0
- 机器人类型:so_follower
- 总片段数 (episodes):100
- 总帧数 (frames):51,387
- 总任务数 (tasks):1
- 块大小 (chunksize):1000
- 数据文件大小:约 100 MB
- 视频文件大小:约 200 MB
- 帧率 (fps):30
- 数据路径格式:
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet - 视频路径格式:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4
数据划分
- 训练集 (train):全部 100 个片段 (索引 0 至 99)
数据特征
| 特征名 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
| action | float32 | (6,) | 6 个关节动作:shoulder_pan.pos, shoulder_lift.pos, elbow_flex.pos, wrist_flex.pos, wrist_roll.pos, gripper.pos |
| observation.state | float32 | (6,) | 6 个关节状态(与 action 的关节名称相同) |
| observation.images.front | 视频 (AV1, yuv420p) | (480, 640, 3) | 前摄像头图像,帧率 30 fps |
| observation.images.right | 视频 (AV1, yuv420p) | (480, 640, 3) | 右摄像头图像,帧率 30 fps |
| observation.images.wrist | 视频 (AV1, yuv420p) | (480, 640, 3) | 腕部摄像头图像,帧率 30 fps |
| 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 |
视频详情
- 所有摄像头图像高度为 480 像素,宽度为 640 像素,3 个通道。
- 视频编码:AV1,像素格式 yuv420p。
- 帧率:30 fps。
- 无音频。
引用信息
- 目前 BibTeX 引用信息为 [More Information Needed]。



