遇见数据集

piper_food100

收藏
Hugging Face2026-07-10 更新2026-07-10 收录
官方服务:

资源简介:

该数据集是一个使用LeRobot工具创建的机器人数据集,专门针对piper_dataset机器人类型。它包含100个episodes,总计120,063帧,分为训练集(索引范围0:100)。数据以30帧/秒的帧率采集,包括300个视频。数据集结构包括时间戳、帧索引、episode索引、任务索引等元数据,以及观察和动作特征。观察特征包括来自三个摄像头(cam_global、cam_gripper、cam_local)的图像数据,每个图像的分辨率为480x640,3通道,并采用libx264编码和yuv420p像素格式。状态观察包含7个关节和夹爪的状态值,动作数据也对应7个维度。数据集以Apache 2.0许可证发布,适用于机器人任务。

This dataset is a robotics dataset created using the LeRobot tool, specifically designed for the piper_dataset robot type. It contains 100 episodes, totaling 120,063 frames, divided into a training set (index range 0:100). The data was collected at a frame rate of 30 frames per second and includes 300 videos. The dataset structure includes metadata such as timestamps, frame indices, episode indices, task indices, as well as observation and action features. Observation features include image data from three cameras (cam_global, cam_gripper, cam_local), each image with a resolution of 480x640, 3 channels, encoded with libx264 and yuv420p pixel format. State observations contain values for 7 joints and the gripper, and action data also corresponds to 7 dimensions. The dataset is released under the Apache 2.0 license and is suitable for robotics tasks.

提供机构:
vis22
创建时间:
2026-07-10
原始信息汇总

数据集摘要

  • 名称:piper_food100
  • 许可证:Apache-2.0
  • 任务类别:机器人技术(robotics)
  • 创建工具:LeRobot

数据集结构

属性
机器人类型 piper_dataset
总片段数 100
总帧数 120,063
总任务数 1
总视频数 300
总数据块数 1
数据块大小 1,000
帧率(FPS) 30
训练集划分 0:100(全部)

数据存储路径

  • 数据文件data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet
  • 视频文件videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4

数据特征

特征字段 数据类型 形状/尺寸 说明
timestamp float32 [1] 时间戳
frame_index int64 [1] 帧索引
episode_index int64 [1] 片段索引
index int64 [1] 索引
task_index int64 [1] 任务索引
observation.images.cam_global video [480, 640, 3] 全局摄像头图像(H.264编码,yuv420p)
observation.images.cam_gripper video [480, 640, 3] 夹爪摄像头图像(H.264编码,yuv420p)
observation.images.cam_local video [480, 640, 3] 局部摄像头图像(H.264编码,yuv420p)
observation.state float32 [7] 机器人状态(6个关节 + 1个夹爪)
action float32 [7] 动作指令(6个关节 + 1个夹爪)

引用信息

  • 论文:暂缺(More Information Needed)
二维码
社区交流群
二维码
科研交流群
商业服务