20s
Datasets
All datasets matching “20s”hq-20snine-source-hand-data-review-videos-20s
训练数据审核视频：MANO GT 优先
最新用户要求：有官方 MANO GT 的数据使用 MANO GT；没有 MANO GT 的数据使用我们自己的 V9-P3R2。文件名和 video_index.csv 明确标注来源。
MANO_GT：数据集发布的 MANO 标注，包括官方拟合标注（official fitted annotation），不是我们另行拟合的结果，也不代表标注没有误差。
P3R2_PREDICTED：我们的实际管线输出，不是 GT。EgoDex 虽有其他手部标签，当前素材没有原生 MANO；用预测 mesh 不等于验证了原始标签正确。EPIC 同样使用预测 MANO。
左右栏使用同一来源、同帧 MANO；保留彩色场景点云、手轨迹和相机。复用已验证渲染，没有把模型错误或 GT 异常抹掉。部分既有视频页脚写 native / official fitted annotation，与文件名 MANO_GT 对应。
本文件夹共 407 个条目文件，来自 398 个独立上下文；别名不能重复计作独立交互时长。
数据集
MANO 来源… See the full description on the dataset page: https://huggingface.co/datasets/yangzijing/nine-source-hand-data-review-videos-20s.everyayah_curated_1s_20s_balancedso101_grab_Gcube_20sec_200ep_merged_102725
so101_grab_Gcube_20sec_200ep_merged_102725
This dataset was generated using phosphobot.
This dataset contains a series of episodes recorded with a robot and multiple cameras. It can be directly used to train a policy using imitation learning. It's compatible with LeRobot.
To get started in robotics, get your own phospho starter pack..
everyayah_curated_1s_20sso101_grab_dice_20sec_50ep_pt2_101425This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "so101_follower",
"total_episodes": 50,
"total_frames": 29949,
"total_tasks": 1,
"total_videos": 100,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:50"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/qm30631122/so101_grab_dice_20sec_50ep_pt2_101425.