Skip to content
Lucie Bierent edited this page Jan 20, 2026 · 17 revisions

1. Introduction and Background

2. Project Structure

3. Useful Links

📘 Annotation Manual: /orcd/data/satra/002/datasets/SAILS/data4analysis/Manuals & Measure Info/Video Rating Manual/SAILS_Home_Video_Rating_Manual_11.4.25.pdf

📄 ASD Status Labels: /orcd/data/satra/002/datasets/SAILS/data4analysis/ASD_Status.xlsx

📊 BIDS Dataset README: /orcd/scratch/bcs/001/sensein/sails/BIDS_data/final_bids-dataset/README

COI Tracking Data: /orcd/data/satra/002/datasets/SAILS/SFARI_Database/COI_tracking

Name of the participants : /orcd/scratch/bcs/001/sensein/sails/SAILS_names.xlsx

4. SAILS Data

IMPORTANT NOTE : In the dataset, we couldn't extract the age/timepoint of the kid some videos. We are still considering them to train/finetune/test our annotator models, but we should drop these videos for further analysis (i.e model focusing on the ASD diagnose based on the evolution through time of the annotated behaviors of the kid). Indeed, droping them would represent letting go of 30% of our annotated data (992/3411).

4.1 Phase III Videos

Filepath: /orcd/data/satra/002/datasets/SAILS/Phase_III_Videos/Videos_from_external: Raw video data - should not be touched.
Filepath: /orcd/data/satra/002/datasets/SAILS/Phase_III_Videos/Videos_from_external_standardized: Converted MP4 data of all videos to be analyzed.
Filepath : /orcd/data/satra/002/datasets/SAILS/Phase_III_Videos/Videos_from_external_categorized: Videos organized in subfolders on face-body visibility criteria. It only contains a subpart of the dataset, including videos that fall within the categories defined below: :

Face-body visibility quality Face and body visibility score (from annotations)
perfect face ≥ 9 & body ≥ 9
good 6 ≤ face ≤ 8 & 6 ≤ body ≤ 8
poor face < 4 & body < 4

Data should be copied to a scratch folder for analysis; the /orcd/data/satra/002/datasets/SAILS folder should not be written to.

4.2 BIDS-Formatted videos

We work exclusively with BIDS-formatted data. Find the converted videos and audios here: /orcd/scratch/bcs/001/sensein/sails/BIDS_data/final_bids-dataset

For details on structure, see the dataset README at: /orcd/scratch/bcs/001/sensein/sails/BIDS_data/final_bids-dataset/README

4.3 Annotation for videos

The home videos were manually anotated, giving information on the subject and the video, very relevant for further analysis (please check the manual).

Path Description
/orcd/data/satra/002/datasets/SAILS/data4analysis/Video Rating Data Directory containing annotations spreadsheets. Use the most recent file
/orcd/scratch/bcs/001/sensein/sails/BIDS_data/anotated_processed.csv Mapping between annotations and corresponding BIDS video paths.
/orcd/scratch/bcs/001/sensein/sails/BIDS_data/unanotated_videos.csv List of videos without annotations.
/orcd/scratch/bcs/001/sensein/sails/BIDS_data/annotated_videos_missing.csv Annotated videos missing BIDS conversion.

5. Scratch Folder

Filepath: /orcd/scratch/bcs/001/sensein/sails: Move files from personal scratch folder here for group access

6. Labels

Participant diagnosis information: /orcd/data/satra/002/datasets/SAILS/data4analysis/ASD_Status.xlsx : contains the ID and diagnosis (ASD or Non-ASD) for all 151 participants.

7. Current Status

Child of Interest Identification

Tracking data is being uploaded as it is done. Currently IDS beginning with A-C are uploaded to engaging at /orcd/data/satra/002/datasets/SAILS/SFARI_Database/COI_tracking