Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/hdf5/hdf5.py", line 49, in _split_generators
                  import h5py
              ModuleNotFoundError: No module named 'h5py'
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 71, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

UMI Raw Chase

这是 Chase 操作任务的原始 UMI 双手示范数据集。数据以 HDF5 文件保存,每个 episode_XX.hdf5 对应一条成功示范轨迹。

数据集概览

  • 共 78 个 episodes、40,064 帧,大小约 850 MB。
  • 采集频率为 30 Hz。
  • 每条轨迹包含左右手腕 RGB 图像、左右手 FastUMI 位姿、左右夹爪输入和时间戳。
  • RGB 图像以 JPEG 压缩后存储,记录的图像尺寸属性为 320。
  • 所有已收录 episode 的 success 属性均为 true

目录与命名

目录 Episodes 帧数 说明
local_chase/ 10 4,597 基础 Chase 数据
local_chase_disturb/ 10 4,536 画面中存在不同颜色、形状的干扰物体
local_chase_fall/ 5 3,170 目标物体抓取后跌落,再次抓取
local_chase_fallre/ 3 2,139 放入收纳盒失败后,再次放入
local_chase_li/ 10 4,802 由另一位采集者采集
local_chase_offset/ 20 10,675 目标物体在上、下、左、右方向存在位置偏置
local_chase_offsetre/ 10 4,383 收纳盒在上、下、左、右方向存在位置偏置
mobile_chase_li/ 10 5,762 mobile 模式,由另一位采集者采集

命名后缀含义:

  • offset:目标物体存在上、下、左、右位置偏置。
  • fall:目标物体抓取后跌落,需要二次抓取。
  • disturb:画面中有不同颜色、形状的干扰物体。
  • li:由另一位采集者采集。
  • re:场景涉及收纳盒;例如 fallre 表示第一次未成功放入收纳盒后再次放入, offsetre 表示收纳盒存在位置偏置。

HDF5 结构

典型 episode 包含:

observations/
  images/
    wrist_left_rgb          # JPEG 压缩的左手腕 RGB 帧
    wrist_left_rgb_len      # 每帧 JPEG 数据的有效字节数
    wrist_right_rgb         # JPEG 压缩的右手腕 RGB 帧
    wrist_right_rgb_len     # 每帧 JPEG 数据的有效字节数
raw_input/
  clamp_left                # 左夹爪输入
  clamp_right               # 右夹爪输入
  fastumi_pose_left         # 左手 FastUMI 7 维位姿
  fastumi_pose_right        # 右手 FastUMI 7 维位姿
timestamps                  # 各帧时间戳

根节点属性还包含 taskbatchcollectornum_stepsrecord_freq_hzimage_compressionimage_sizesuccess 等采集元数据。

读取压缩图像时,应先使用对应的 *_len 截取该帧的有效字节,再进行 JPEG 解码。

快速读取

import cv2
import h5py
import numpy as np

with h5py.File("local_chase/episode_01.hdf5", "r") as f:
    timestamps = f["timestamps"][:]
    left_pose = f["raw_input/fastumi_pose_left"][:]

    encoded = f["observations/images/wrist_left_rgb"][0]
    encoded_len = f["observations/images/wrist_left_rgb_len"][0]
    left_image = cv2.imdecode(
        np.asarray(encoded[:encoded_len], dtype=np.uint8),
        cv2.IMREAD_COLOR,
    )

使用说明

该仓库保存未经转换的原始采集数据。训练前请根据任务需要完成图像解码、轨迹对齐、 动作定义确认及数据格式转换。FastUMI 7 维位姿的各维语义和夹爪数值范围应以对应的 采集程序为准。

Downloads last month
134