fold-towel-0701
收藏资源简介：
该数据集是一个使用LeRobot工具创建的机器人操作数据集，专注于机器人任务，特别是双TacCap抓取器（bi_taccap_gripper）的操作。它包含20个完整的情节（episodes），总计26876帧数据，帧率为30fps。数据以训练集形式组织，并分为多个块（chunks）。数据集内容包括机器人的动作和观测：动作涉及左右机械臂的TCP（工具中心点）位置、旋转和夹爪位置；观测包括状态信息（结构与动作类似）以及多种视觉数据，具体为左触觉左传感器、左触觉右传感器、左腕部摄像头、右触觉左传感器、右触觉右传感器和右腕部摄像头的视频数据，这些视频具有不同的分辨率和编码格式。此外，数据集还包含时间戳、帧索引、情节索引等元数据。总数据文件大小约为100MB，视频文件大小约为200MB。
This dataset is a robotic manipulation dataset created using the LeRobot tool, focusing on robotic manipulation tasks with a particular emphasis on operations utilizing the bi_taccap_gripper. It contains 20 complete episodes, totaling 26876 frames at a frame rate of 30 fps. The dataset is organized as a training dataset and split into multiple chunks. The dataset content includes robot actions and observations: Actions encompass the TCP (Tool Center Point) positions, orientations, and gripper positions of both the left and right robotic arms; Observations comprise state information with a structure analogous to that of the actions, as well as multiple visual modalities, specifically video streams from the left tactile left sensor, left tactile right sensor, left wrist camera, right tactile left sensor, right tactile right sensor, and right wrist camera, each with distinct resolutions and encoding formats. Additionally, the dataset includes metadata such as timestamps, frame indices, and episode indices. The total size of the non-visual data files is approximately 100 MB, while the total size of the video files is roughly 200 MB.
数据集概述：fold-towel-0701
- 许可证: Apache-2.0
- 任务类别: 机器人学 (Robotics)
- 创建工具: LeRobot
数据集详情
| 属性 | 值 |
|---|---|
| 机器人类型 | bi_taccap_gripper |
| 总集数 | 20 |
| 总帧数 | 26,876 |
| 任务数量 | 1 |
| 块大小 | 1000 |
| 数据文件大小 | 100 MB |
| 视频文件大小 | 200 MB |
| 帧率 | 30 FPS |
| 数据分割 | 训练集: 0–20 集 |
| 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 |
特征结构
动作 (action)
- 数据类型: float32
- 形状: [20]
- 维度名称:
left_tcp.x,left_tcp.y,left_tcp.z,left_tcp.r1–left_tcp.r6,left_gripper.pos,right_tcp.x,right_tcp.y,right_tcp.z,right_tcp.r1–right_tcp.r6,right_gripper.pos
观察状态 (observation.state)
- 数据类型: float32
- 形状: [20]
- 维度名称: 同
action
观察图像 (observation.images)
| 图像键 | 形状 (高度×宽度×通道) | 视频编解码 | 帧率 | 深度映射 |
|---|---|---|---|---|
| left_tactile_left | 400×700×3 | h264 | 30 | 否 |
| left_tactile_right | 400×700×3 | h264 | 30 | 否 |
| left_wrist | 480×640×3 | h264 | 30 | 否 |
| right_tactile_left | 400×700×3 | h264 | 30 | 否 |
| right_tactile_right | 400×700×3 | h264 | 30 | 否 |
| right_wrist | 480×640×3 | h264 | 30 | 否 |
其他特征
- timestamp: float32, 形状 [1]
- frame_index: int64, 形状 [1]
- episode_index: int64, 形状 [1]
- index: int64, 形状 [1]
- task_index: int64, 形状 [1]
引用信息
bibtex @misc{xense-taccap-lerobot, author = {XenseRobotics Team}, title = {LeRobot-Xense: LeRobot with Xense Tactile Robotics Support}, howpublished = {url{https://github.com/Vertax42/xense-taccap-lerobot}}, year = {2026} }
