sim_grasp_v9_cartesian
收藏资源简介:
该数据集是一个用于机器人抓取任务的模拟数据集,使用LeRobot工具创建。数据集包含168个episodes,总计57900帧,帧率为50fps。数据特征包括:观察图像(来自相机cam0,分辨率为972x1296,3通道),动作数据(11维浮点数组,涵盖位移、旋转和夹爪控制),立方体姿态(7维浮点数组,表示位置和四元数旋转),以及时间戳、帧索引、episode索引等元数据。机器人类型为openarm_gripette_sim,数据以parquet文件格式存储,视频以mp4格式存储。数据集主要用于训练和评估机器人抓取算法,支持计算机视觉和强化学习应用。
This dataset is a simulation dataset for robot grasping tasks, created using the LeRobot tool. It contains 168 episodes, totaling 57,900 frames at a frame rate of 50fps. The data features include: observation images (from camera cam0, with a resolution of 972x1296 and 3 channels), action data (11-dimensional floating-point arrays covering displacement, rotation, and gripper control), cube poses (7-dimensional floating-point arrays representing position and quaternion rotation), as well as metadata such as timestamps, frame indices, and episode indices. The robot type is openarm_gripette_sim, with data stored in parquet file format and videos in mp4 format. The dataset is primarily used for training and evaluating robot grasping algorithms, supporting computer vision and reinforcement learning applications.
数据集概述:sim_grasp_v9_cartesian
- 许可证: Apache-2.0
- 任务类别: 机器人学
- 标签: LeRobot
- 创建工具: LeRobot
数据集规模与结构
- 代码库版本: v3.0
- 机器人类型: openarm_gripette_sim
- 总片段数: 168
- 总帧数: 57900
- 总任务数: 1
- 块大小: 1000
- 数据文件大小: 100 MB
- 视频文件大小: 200 MB
- 帧率 (FPS): 50
- 数据分割: 训练集包含全部168个片段(索引0至167)
数据文件格式
- 数据路径:
data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet - 视频路径:
videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4
特征详解
| 特征名称 | 数据类型 | 形状 | 说明 |
|---|---|---|---|
observation.images.cam0 |
视频 | [3, 972, 1296] | 相机图像,尺寸972x1296,AV1编码,50 FPS,3通道,非深度图 |
action |
float32 | [11] | 动作向量:dx, dy, dz, dr6d_0~dr6d_5, proximal, distal |
cube_pose |
float32 | [7] | 立方体位姿:x, y, z, qw, qx, qy, qz |
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 |
observation.state |
float32 | [2] | 机器人状态:proximal, distal |
可视化
可通过 LeRobot 可视化工具 直接浏览该数据集。



