遇见数据集

cell1_20260515_light-green-short-sleeve-tshirt_folding_unfolding_fault_episode-1-2

收藏
Hugging Face2026-05-21 更新2026-05-21 收录
官方服务:

资源简介:

该数据集由LeRobot创建,专注于机器人学领域,特别是双机械臂(型号为starpilot_yam_gripper)的演示任务。数据集包含折叠和展开任务的相关数据,旨在支持机器人学习研究。它包含2个完整的episodes,总计1450帧,覆盖1个具体任务。数据以parquet文件格式存储,视频文件为MP4格式,帧率为30fps。数据集的特征丰富,包括动作数据(14维浮点数,表示两个机械臂的x、y、z位置、rx、ry、rz旋转以及夹爪宽度)、观察状态数据(32维浮点数,包括机械臂的编码器角度、IMU加速度和陀螺仪数据、姿态四元数、夹爪开合状态和距离),以及来自多个摄像头的RGB图像和深度图像视频(摄像头包括左腕0号、右腕0号、基座0号和基座1号,分辨率分别为480x640和768x1024)。此外,还提供时间戳、帧索引、episode索引、任务索引等元数据,便于数据分析和模型训练。数据集采用Apache 2.0许可证。

This dataset was developed by LeRobot, focusing on the field of robotics, particularly demonstration tasks for dual robotic arms with the model starpilot_yam_gripper. It contains data related to folding and unfolding tasks, aiming to support robotic learning research. The dataset includes 2 complete episodes, totaling 1450 frames and covering 1 specific task. The data is stored in Parquet file format, with video files in MP4 format at a frame rate of 30 fps. The dataset has rich multimodal features, including: 1. Action data: 14-dimensional floating-point numbers, representing the x, y, z positions, rx, ry, rz rotations and gripper width of the two robotic arms; 2. Observation state data: 32-dimensional floating-point numbers, including encoder angles of the robotic arms, IMU acceleration and gyroscope data, attitude quaternions, gripper opening/closing status and distance measurements; 3. RGB and depth image videos from multiple cameras: left wrist 0, right wrist 0, base 0 and base 1, with resolutions of 480×640 and 768×1024 respectively. In addition, metadata such as timestamps, frame indices, episode indices and task indices are provided to facilitate data analysis and model training. The dataset is licensed under Apache 2.0.

提供机构:
nodogoro
创建时间:
2026-05-21
原始信息汇总

数据集概述

  • 名称: cell1_20260515_light-green-short-sleeve-tshirt_folding_unfolding_fault_episode-1-2
  • 许可证: Apache-2.0
  • 任务类型: 机器人技术(Robotics)
  • 创建工具: 使用 LeRobot 创建

数据集结构

  • 机器人类型: starpilot_yam_gripper
  • 总片段数: 2
  • 总帧数: 1450
  • 总任务数: 1
  • 数据分块大小: 1000
  • 数据文件大小: 约 100 MB
  • 视频文件大小: 约 200 MB
  • 帧率: 30 FPS
  • 数据集划分: 仅包含训练集(索引 0:2)

特征信息

特征名称 数据类型 形状 说明
action float32 [14] 包含双臂位置、旋转和夹爪宽度(14维)
observation.state float32 [32] 包含双臂编码器值、IMU数据、位姿和夹爪状态(32维)
observation.images.left_wrist_0_camera_rgb_image 视频 [480, 640, 3] 左腕RGB相机图像(H.264编码,30 FPS)
observation.images.left_wrist_0_camera_rgb_image_depth 视频 [480, 640, 3] 左腕RGB深度图(H.264编码,30 FPS)
observation.images.right_wrist_0_camera_rgb_image 视频 [480, 640, 3] 右腕RGB相机图像(H.264编码,30 FPS)
observation.images.right_wrist_0_camera_rgb_image_depth 视频 [480, 640, 3] 右腕RGB深度图(H.264编码,30 FPS)
observation.images.base_0_camera_rgb_image 视频 [480, 640, 3] 基座相机0 RGB图像(H.264编码,30 FPS)
observation.images.base_1_camera_rgb_image 视频 [768, 1024, 3] 基座相机1 RGB图像(H.264编码,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

可视化

可通过 LeRobot 可视化空间 查看数据集样例。

二维码
社区交流群
二维码
科研交流群
商业服务