遇见数据集

2026-06-10_shakeitup_aluboxvf_5_6kHz_band

收藏
Hugging Face2026-06-10 更新2026-06-10 收录
官方服务:

资源简介:

该数据集是一个基于LeRobot平台创建的机器人技术数据集,专为机器人控制和模仿学习研究设计。数据集包含50个训练片段,总计23955帧,帧率为30fps。数据特征涵盖:动作空间(包括6自由度关节位置,具体为肩部平移、肩部提升、肘部弯曲、腕部弯曲、腕部旋转和夹爪位置)、观测状态(与动作相同的6自由度关节位置)、图像观测(包括顶部摄像头和腕部摄像头的480x640分辨率RGB视频,以及5.6kHz触觉频谱图的224x224分辨率视频)。此外,数据集还提供时间戳、帧索引、片段索引、任务索引等元数据。机器人类型为so101_follower_dragon_multitactile,数据以parquet文件和MP4视频格式存储,总数据量约为300MB(其中视频部分占200MB)。该数据集适用于机器人技术领域的多种研究任务,如控制算法开发和模仿学习模型训练。

This robotic dataset is developed on the LeRobot platform, tailored for robotic control and imitation learning research. The dataset consists of 50 training episodes, with a total of 23955 frames at a frame rate of 30fps. The data features include: action space (6-degree-of-freedom joint positions, specifically shoulder translation, shoulder lift, elbow flexion, wrist flexion, wrist rotation, and gripper position), observation states (same 6-degree-of-freedom joint positions as the action space), and visual observations (480x640 resolution RGB videos captured by both the top camera and wrist camera, as well as 224x224 resolution videos of 5.6kHz tactile spectrograms). In addition, the dataset also provides metadata such as timestamps, frame indices, episode indices, and task indices. The robotic platform type is so101_follower_dragon_multitactile. The data is stored in Parquet files and MP4 video formats, with a total size of approximately 300MB (of which the video portion accounts for 200MB). This dataset is applicable to a variety of research tasks in the field of robotics, such as control algorithm development and imitation learning model training.

提供机构:
jogarulfop
创建时间:
2026-06-10
原始信息汇总

数据集概述

  • 数据集名称: jogarulfop/2026-06-10_shakeitup_aluboxvf_5_6kHz_band
  • 许可证: Apache-2.0
  • 任务类别: 机器人学 (robotics)
  • 创建工具: 使用 LeRobot 创建

数据集结构

  • 总片段数: 50
  • 总帧数: 23955
  • 总任务数: 1
  • 帧率: 30 FPS
  • 机器人类型: so101_follower_dragon_multitactile
  • 数据分割: 仅包含训练集 (0:50)
  • 数据文件大小: 约 100 MB
  • 视频文件大小: 约 200 MB

特征

特征名称 数据类型 形状 说明
action float32 [6] 机器人动作,包含6个关节位置 (shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper)
observation.state float32 [6] 机器人观察状态,与动作空间相同 (6个关节位置)
observation.images.top video [480, 640, 3] 顶部摄像头视频,分辨率480x640,30 FPS,AV1编码
observation.images.wrist video [480, 640, 3] 腕部摄像头视频,分辨率480x640,30 FPS,AV1编码
observation.images.5_6kHz_tactile_spectrogram video [224, 224, 3] 5.6 kHz触觉频谱图视频,分辨率224x224,30 FPS,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] 全局索引
task_index int64 [1] 任务索引

其他信息

  • 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
  • 代码库版本: v3.0
  • 引用: 暂无可用 BibTeX 引用信息。
二维码
社区交流群
二维码
科研交流群
商业服务