This dataset is derived from EPIC-KITCHENS-100, which is CC BY-NC 4.0. That term carries into everything built from it, so this whole repository — the Parquet tables and the arm-removed video alike — is available for non-commercial use only. It is the one HuRo subset whose tables are non-commercial, and the source is the reason.

No unmodified EPIC-KITCHENS frames and no robot-rendered frames are included.

Your username, email address and the fields below are collected solely to administer access to this dataset and to evidence acceptance of these terms. They may be provided in aggregate to the licensors of the material for the purpose of verifying compliance.

HuRo-EPIC-KITCHENS

The EPIC-KITCHENS-100 share of the HuRo dataset: 15,935 segments / 2,456,378 frames across 636 clips, 22.7 hours at 30 fps. Each segment holds the frames with the human arms inpainted out, per-frame hand and camera annotation, a language instruction, and the Allex joint trajectory retargeting produced from them.

This is the robot-independent half of HuRo. No robot is rendered into the frames: you retarget to the robot you care about and render it yourself, which is stages 8 to 10. The expensive stages 1 to 7 — hand reconstruction, camera estimation, narration, inpainting — are already done here and never run again.

Produced by the HuRo robotization pipeline. Paper: arXiv:2609.10706 · Project page: 3587jjh.github.io/HuRo

What's in this repository

The repository name says which corpus this came from, not what is inside. What is here:


Human annotation tables	✅ `clips_chunked/original/annot/` — one Parquet table per segment, robot-independent
Arm-removed video	✅ `clips_chunked/original/video/` — one `*_inpainted.mp4` per segment, the frames with the human arms painted out
Robot trajectory tables	✅ ALLEX — `clips_chunked/allex/annot/`
Unmodified source frames	❌ never — download EPIC-KITCHENS-100 yourself, and `tools/prepare_segments.py` cuts it to these segments
Robot-rendered frames	❌ never — the ALLEX meshes are evaluation-only. Stage 9 renders them on your machine
Robot trees are keyed by robot name, so a second robot arrives as a sibling of `allex/` and
nothing above it changes. ALLEX is the only one published so far.

Licences differ by file class — see Licence.

What you can do with this

EPIC-KITCHENS-100 is CC BY-NC 4.0, which permits sharing adaptations, so the arm-removed video ships here and the annotation half of the pipeline never has to be rerun. 15.2 GiB in total: 3.1 GiB human annotations, 11.4 GiB arm-removed video, 1.7 GiB Allex trajectory tables. 15,901 of the 15,935 segments carry an Allex trajectory (99.8%); the other 34 ship their human annotations alone.

No frames ship unmodified, and no frames ship with a robot in them. Every video file here is a segment with the human arms inpainted out and nothing put back in their place.

The original pixels are not included. Download EPIC-KITCHENS-100 yourself and cut it to our segment boundaries.
The Allex-rendered pixels are not included either. pipeline/stage9_robot_overlay.py reproduces them from what is here, and stage10_lerobot_convert.py packs the result as a LeRobot V2.0 dataset. Both need a GPU; nothing else in the pipeline has to run.

The human annotations are robot-independent: stages 1 to 7 of the pipeline never see a robot.

You want	You need	Stages you run
Hand and camera annotations over arm-removed video	`original/`	none
Allex joint trajectories, to train on states or to check ours	`allex/annot/`	none
Robotize onto your own robot	`original/`	8 → 9 → 10
The Allex LeRobot dataset the paper trains on	`original/` + `allex/annot/`	9 → 10

Row three is the point of this release. Because the arm-removed video ships alongside the annotations, retargeting to a different robot skips the whole annotation half of the pipeline and the inpainting — you go straight to IK and rendering. Adding a robot needs one YAML file and a URDF, no code change. See Robotizing onto your own robot.

Row four needs the Allex meshes, which come from the ALLEX model repo and carry their own licence — see ALLEX below. Nothing in this repository depends on them.

Read this first

Access is gated. Accept the terms on the dataset page and access is granted automatically. The acceptance is recorded; see Licence for what the terms commit you to.
Arm-removed video is included; original video and robot renders are not. EPIC-KITCHENS-100 is CC BY-NC 4.0, which permits sharing adaptations non-commercially with attribution — so the inpainted frames ship. The unaltered source frames do not, and neither does any frame with a robot composited into it.
Non-commercial only, and doubly so — EPIC-KITCHENS is NC, and the pipeline's own components (HaWoR, ProPainter, MANO) are non-commercial or research-only.
Every frame is modified from the original EPIC-KITCHENS material: undistorted and with the arms inpainted out. Stated here as CC BY-NC 4.0 requires.
States and actions are pipeline estimates, from monocular hand reconstruction and IK, not motion capture. Pre-training corpus, not an evaluation benchmark.
Compute your own normalisation statistics. Stage 10 leaves meta/stats.json out of the LeRobot dataset it builds, because the statistics depend on the corpus and resolution mix you train on. Follow the format of lerobot/pusht.

clips_chunked/
├── original/annot/<clip_id>/<start>_<end>.parquet         human annotations, stages 2-7
├── original/video/<clip_id>/<start>_<end>_inpainted.mp4   the segment's frames, arms removed (stage 7)
├── allex/annot/<clip_id>/<start>_<end>.parquet            Allex joint trajectory, stage 8
├── original/video/<clip_id>.done                           stage 6 marker
├── original/annot/<clip_id>.done                           stage 7 marker
└── allex/annot/<clip_id>.done                              stage 8 marker

segments.csv                                           clip_id, start, end, fps, resolution, language

Three .done markers per clip, which is what the pipeline reads: stage 7 checks stage 6's original/video/<clip>.done, stage 8 is satisfied by original/annot/<clip>.done, and stage 9 requires its robot's stage-8 marker, allex/annot/<clip>.done. There is deliberately no stage-9 marker — stage 9 skips every clip whose marker already exists (stage9_robot_overlay.py:368-372), so shipping one would make your own render silently produce nothing.

allex/annot/ carries the full 37-column schema, including overlay_valid — the per-frame flag saying the Allex projection was usable in that frame when we rendered it. Use it to drop bad frames before you spend GPU time rendering.

The clips_chunked/ tree is the same layout the HuRo-Ego4D repository ships, plus the modified video that Ego4D's licence forbids there. Every column is documented under Columns below, and in examples/README.md.

Rebuilding things yourself

Cutting your own segments from the source corpus, retargeting onto a different robot, rendering it into the frames, and packing the result as a LeRobot dataset are all documented in REBUILDING.md in this repository. None of it is needed to use the dataset as published.

Columns

One row per video frame, in the 37-column schema shared by every HuRo release. Both tables carry all 37 columns; they differ in which are filled. Read a row with decode_row from common/io.py — arm_mask is bit-packed and absent per-hand arrays are stored as NaN lists — or run the repo's reader:

python examples/read_parquet.py clips_chunked

`original/annot/<clip_id>/<start>_<end>.parquet` — robot-independent

Column	Type	Meaning
`clip_id`, `frame_id`	string, int32	EPIC-KITCHENS video id; frame index, restarting at 0 per segment
`height`, `width`	int32	the rectified frame size the annotations were made at
`intr_model`	string	which calibration produced the raw intrinsics — `droidcalib` on 15,478 of the 15,935 segments, `anycalib` on the remaining 457
`fx`, `fy`, `cx`, `cy`, `xi`	float64	the camera model as recorded. `xi` is non-zero on 15,592 of the 15,935 segments (it ranges from −0.0068 to 1.2625), so those frames are rectified before the annotations are made
`pinhole_fx`, `pinhole_fy`, `pinhole_cx`, `pinhole_cy`	float64	the rectified pinhole every geometry column here is expressed against. Project keypoints with these, not with `fx`/`fy`
`cam_pose`	float32[4,4]	cam-to-world, OpenCV, metric-scaled and gravity-aligned
`arm_mask`	binary, bit-packed (H,W) bool	the human-arm mask stage 7 inpainted. Filled on every row, and the reason you can skip stages 1–6
`hands[]`	list of struct	one entry per detected hand — see below
`narr`	struct	`think`, `left`, `right`, `bimanual` — the VLM's per-segment narration
`language`	string	the merged instruction, one per segment
`n_person_det`	int32	null in this subset

hands[] fields: box float32[4] [x1,y1,x2,y2] in the rectified frame, conf, side int8 (0 left, 1 right), side_conf, kpts3d float32[21,3] in MediaPipe order and camera frame, wrist_rot float32[3] axis-angle — null where stage 5 could not resolve the wrist orientation — finger_rot float32[15,3] axis-angle, and hand_mask, which is null on every row in every HuRo subset: the per-hand masks were transient inputs to inpainting and were never persisted.

The stage-8 columns listed below are present but null in this table, as are n_person_det and overlay_valid.

`allex/annot/<clip_id>/<start>_<end>.parquet` — Allex only

Skip this table if you are targeting your own robot; stage 8 regenerates its equivalent from original/annot.

Column	Type	Meaning
`state_qpos`	float32[48]	Allex joint angles, ordered by the `state_qpos_joint_names` metadata
`state_eef_left`, `state_eef_right`	float32	wrist position (3) + rotation 6D (6) + hand joint angles
`state_mask_left`, `state_mask_right`	bool	false where that side's hand is missing
`cam_pose_base`	float32[4,4]	cam-to-robot-base, OpenCV
`retarget_cost`, `err_ddq`, `err_tip_mm_`, `err_palm_mm_`, `err_local_dir_*`, `err_cam_pos_mm`, `err_cam_rot_deg`	float32	IK diagnostics; stage 10 filters episodes on the `err_*` columns
`overlay_valid`	bool	stage 9's per-frame flag saying the Allex projection was usable. Use it to drop bad frames before you spend GPU time rendering

Schema metadata, readable via pq.read_table(...).schema.metadata and attached to the allex tables only: robot_name, state_qpos_joint_names, state_eef_layout (per_side:wrist_pos3_rot6d6_hand_joints), state_eef_left_hand_joint_names, state_eef_right_hand_joint_names, state_eef_frame (robot_base_frame), cam_pose_convention (opencv_cam_to_base).

Positions in metres, angles in radians. Robot base frame is x forward, y left, z up; camera axes are OpenCV.

Other HuRo subsets

Subset	Repo	Pixels included?
EPIC-KITCHENS-100	this repo	arm-removed frames only — no original pixels, no robot rendered
Egocentric-10K	`Joocjun/HuRo-Egocentric10K`	arm-removed frames only — no original pixels, no robot rendered
Ego4D	`Joocjun/HuRo-Ego4D`	annotations only — its licence forbids the video; bring your own copy
EgoVerse	`Joocjun/HuRo-EgoVerse`	arm-removed frames only — no original pixels, no robot rendered
EgoDex	not released — CC BY-NC-ND 4.0 forbids sharing adapted material. regenerating it from your own copy is possible with this pipeline, but the recipe document is not published yet	—

Licence

This dataset is not a single licence. Three layers stack, and the narrowest one wins for any given file:

Layer	Sets
The source corpus	what may be redistributed of this subset at all — here CC BY-NC 4.0, which is what makes this whole repository non-commercial
The pipeline	reaches the arm-removed frames, which are ProPainter output (S-Lab 1.0, non-commercial). Whether a tool's licence reaches its output is unsettled; here it changes nothing, because the source already requires non-commercial
The robot model	the ALLEX URDF is BSD 3-Clause and its notice must be kept — see ALLEX below. No ALLEX-licensed pixels are published, so no separate grant applies to any file here

What that means file by file:

File	Licence
Pipeline code (the HuRo repo)	Apache-2.0
Annotation and trajectory tables (`*.parquet`, `segments.csv`, `meta/`)	CC BY-NC 4.0
Arm-removed frames (`*_inpainted.mp4`)	CC BY-NC 4.0

This is the only HuRo subset whose tables are non-commercial because of its source. The other three carry CC BY 4.0 tables; see Other HuRo subsets.

ALLEX

Required attribution, to be kept with any redistribution:

ALLEX robot model and rendered appearance © 2026 WIRobotics Inc. Used with permission. ALLEX model information: https://github.com/wirobotics-rih/allex_model

The ALLEX model repository's BSD 3-Clause covers its MJCF, URDF and USD; the meshes carry a separate, evaluation-only licence. This release publishes nothing derived from the meshes. The Allex tables here come from the URDF's kinematics — joint angles, end-effector poses and an overlay_valid flag — so BSD 3-Clause governs them, and retaining the notice above is the whole of the obligation.

No ALLEX mesh, point cloud, depth map, surface normal or rendered image is published. If you render Allex yourself with stage 9, you obtain the meshes from the ALLEX repository directly and their evaluation-only licence applies to you, not through us.

Source material

The episodes derive from EPIC-KITCHENS-100, © Damen et al., https://epic-kitchens.github.io/, licensed CC BY-NC 4.0 and provided without warranties or conditions of any kind. Every frame is modified from the original material.

The pipeline components retain their own licences — HaWoR (CC BY-NC-ND 4.0), ProPainter (S-Lab 1.0, non-commercial), MANO (research only), and the rest of THIRD_PARTY_NOTICES.md. Review it before redistributing anything derived from this release.

Citation

@misc{jeong2026huro,
      title={HuRo: Robotizing Human Videos for Scalable VLA Pretraining},
      author={Jinho Jeong and Se June Joo and Jaehyun Kang and Dongyun Kim and Yena Kim and Hanjung Kim and Seon Joo Kim},
      year={2026},
      eprint={2609.10706},
      archivePrefix={arXiv},
      primaryClass={cs.RO},
      url={https://arxiv.org/abs/2609.10706},
}

Cite EPIC-KITCHENS-100 as well:

@article{damen2022rescaling,
  title={Rescaling Egocentric Vision: Collection, Pipeline and Challenges for EPIC-KITCHENS-100},
  author={Damen, Dima and Doughty, Hazel and Farinella, Giovanni Maria and Furnari, Antonino and
          Kazakos, Evangelos and Ma, Jian and Moltisanti, Davide and Munro, Jonathan and
          Perrett, Toby and Price, Will and Wray, Michael},
  journal={International Journal of Computer Vision},
  year={2022}
}

Downloads last month: 607

Total file size:

16.3 GB

Collection including Joocjun/HuRo-EPIC-KITCHENS

HuRo

Collection

Robotized egocentric human video. One repo per source dataset, since their licences differ. • 4 items • Updated 1 day ago

Paper for Joocjun/HuRo-EPIC-KITCHENS

HuRo: Robotizing Human Videos for Scalable VLA Pretraining

Paper • 2609.10706 • Published Sep 18 • 24

Accept the terms to access HuRo-EPIC-KITCHENS