record-test2_20260709_132705
收藏资源简介:
该数据集是一个机器人学数据集,使用LeRobot工具创建,专门用于机器人控制和模仿学习任务。数据集包含动作和观测数据,其中动作数据包括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),观测数据包括相同的关节状态以及两个视频流:鸟瞰图(birdsEye)和塔式视图(tower),视频分辨率为480x640像素,帧率为30fps,编码为h264。数据集还包含时间戳、帧索引、episode索引、任务索引等元数据。总共有2个episodes,1068帧,数据以parquet格式分块存储,总数据大小为100MB,视频文件大小为200MB。机器人类型为so_follower,数据集仅包含训练分割。
This dataset is a robotics dataset created using the LeRobot tool, specifically designed for robot control and imitation learning tasks. It includes action and observation data, where action data consists of 6 joint positions (shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, and gripper position), and observation data comprises the same joint states along with two video streams: a birds-eye view (birdsEye) and a tower view (tower). The videos have a resolution of 480x640 pixels, a frame rate of 30fps, and are encoded in h264. The dataset also contains metadata such as timestamps, frame indices, episode indices, and task indices. In total, there are 2 episodes and 1068 frames, with data stored in chunked parquet format, totaling 100MB in size, and video files amounting to 200MB. The robot type is so_follower, and the dataset only includes a training split.
数据集概述
- 数据集名称:
ahmedaali/record-test2_20260709_132705 - 许可证: Apache-2.0
- 任务类别: 机器人学 (Robotics)
- 标签: LeRobot
数据集结构
- 数据类型: 采用 Parquet 和视频文件格式存储
- 配置: 默认配置 (
default),数据文件路径为data/*/*.parquet - 数据量:
- 总集数: 2 集
- 总帧数: 1068 帧
- 总任务数: 1 个任务
- 数据文件大小: 100 MB
- 视频文件大小: 200 MB
特征说明
| 特征名称 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
action |
float32 | [6] | 6 维度动作,包含肩部、肘部、腕部及夹爪位置 |
observation.state |
float32 | [6] | 6 维度状态观测,与动作命名相同 |
observation.images.birdsEye |
视频 (H264) | [480, 640, 3] | 鸟瞰视角图像,分辨率 480x640,30 FPS,RGB 3 通道 |
observation.images.tower |
视频 (H264) | [480, 640, 3] | 塔式视角图像,分辨率 480x640,30 FPS,RGB 3 通道 |
timestamp |
float32 | [1] | 时间戳 |
frame_index |
int64 | [1] | 帧索引 |
episode_index |
int64 | [1] | 集数索引 |
index |
int64 | [1] | 全局索引 |
task_index |
int64 | [1] | 任务索引 |
数据集划分
- 训练集:
0:2(包含全部 2 集数据作为训练集)



