HUI360-Videos-SSUP

Processed SSUP videos for HUI360 — the Cornell SSUP-HRI portion of the HUI360 video dataset, re-encoded to match the HUI360 format.

What is in this repository?

This gated repository contains 29 recordings from the SSUP-HRI dataset (Astor Place and Albee Square deployments, 2022–2023), reprocessed for HUI360:

• video_full_ssup/ — full-resolution episode videos (images_360_full.mp4, 3840×1920)
• video_mini_ssup/ — reduced-resolution episode videos (images_360_mini.mp4, 960×480)

Annotations for these recordings (poses, masks, interaction labels) are freely available in rlorlou/HUI360 and in the annotations/ folder of rlorlou/HUI360-Videos.

Shelf recordings from INRIA (70 recordings) are distributed separately under their own access terms in rlorlou/HUI360-Videos.

Why is access restricted?

SSUP recordings are owned by Cornell's Interaction Research Lab (IRL) and governed by the SSUP-HRI Data Use Agreement. Under clause 5, processed SSUP data cannot be redistributed to users who have not been approved for access to the original dataset.

This repository is therefore gated: we can only grant download access to researchers who already hold a valid SSUP-HRI agreement.

How to get access

You have two options.

Option A — Download processed videos from this repository (recommended if you already have SSUP access)

1. Obtain SSUP-HRI access from Cornell IRL by following the steps on IRL-CT/SSUP-HRI:
   • Obtain IRB approval (or equivalent) from your institution for secondary data use.
   • Sign and return the Data Use Agreement to Fanjun Bu (fb266@cornell.edu).
2. Request access to this repository on Hugging Face (rlorlou/HUI360-Videos-SSUP) using an institutional email, and include proof of your approved SSUP-HRI access (e.g. confirmation email from Cornell IRL).
3. Once approved, download video_full_ssup/ and/or video_mini_ssup/ and merge them into the main HUI360-Videos layout (see Merging with the full dataset below).

Option B — Process the original SSUP videos yourself

If you prefer not to use this repository, you can:

1. Request access to the original SSUP-HRI dataset directly from IRL-CT/SSUP-HRI.
2. Split and re-encode the recordings following the HUI360 episode rules described in rlorlou/HUI360-Videos.

For SSUP recordings (XXX_astor_place_XXX or XXX_albee_square_XXX), episodes are a simple subdivision of the original videos every 5 minutes, except for 2022_09_21_astor_place_recycle, which was accidentally divided into 2-minute sequences.

Merging with the full dataset

After downloading, merge the SSUP directories with the Shelfy portion from rlorlou/HUI360-Videos:

# Merge full-resolution videos
video_full/  ←  video_full_shelfy/  +  video_full_ssup/

# Merge mini-resolution videos
video_mini/  ←  video_mini_shelfy/  +  video_mini_ssup/

The full directory layout, annotation formats, and episode-level file descriptions are documented in the HUI360-Videos README.

SSUP-specific annotation files include interaction_zone_tracking_360.json, which tracks the mobile interaction zone (the trash can) — see the main README for details.

Related resources

Resource	Link
HUI360 project website	hucebot.github.io/hui360
Full HUI360 videos (Shelfy + access info)	rlorlou/HUI360-Videos
Open skeleton annotations	rlorlou/HUI360
SSUP original dataset & agreement	IRL-CT/SSUP-HRI
Annotation & processing pipeline	Interact360
Baselines code	HUI360-Baselines

Citation

If you use these videos, please cite both HUI360 and the original SSUP-HRI dataset:

@INPROCEEDINGS{11556969,
  author={Lorenzo-Louis, Raphael and Amadio, Fabio and Luvison, Bertrand and Ivaldi, Serena},
  booktitle={2026 IEEE 20th International Conference on Automatic Face and Gesture Recognition (FG)},
  title={HUI360 : A 360° Egocentric Dataset and Baselines for Human-Robot Interaction Anticipation},
  year={2026},
  volume={},
  number={},
  pages={1-9},
  doi={10.1109/FG67764.2026.11556969}
}

@inproceedings{BuJ24a,
  author={Bu, Fanjun and Ju, Wendy},
  title={SSUP-HRI: Social Signaling in Urban Public Human-Robot Interaction Dataset},
  booktitle={Social Robotics -- 16th International Conference, ICSR + AI 2024, Odense, Denmark, October 23--26, 2024, Proceedings, Part III},
  series={Lecture Notes in Computer Science},
  volume={15563},
  pages={479--487},
  publisher={Springer},
  year={2024},
  doi={10.1007/978-981-96-3525-2_40}
}