Invariant Versus Context-Specific Representation of Face Shape and Motion in the Face Network

# Invariant Versus Context-Specific Representation of Face Shape and Motion in the Face Network

This repository contains the fMRI dataset associated with the study:

> Hosseini, S. S., & Soto, F. A. (2025). *Invariant versus context-specific representation of face shape and motion in the face network.* Preprint.

## Dataset overview

This dataset includes whole-brain fMRI data from **12 participants** who viewed dynamic synthetic face stimuli in which **facial shape and facial motion were independently manipulated**. The experiment was designed to test whether neural representations of face shape and motion in face-selective cortical regions are invariant or context-specific.

Participants viewed short videos generated from 3D face models, combining:
- 3 facial shapes (identities)
- 3 facial motion patterns (pose-preserving and pose-changing)

The dataset includes:
- Raw functional and anatomical MRI data in **BIDS format**
- Event timing files for all runs
- Functional localizer data for defining face-selective ROIs

All anatomical images have been **defaced** to protect participant privacy.

## Experimental design

The experiment consisted of:
- A behavioral familiarization session
- A face localizer task (faces vs objects)
- A main event-related task with dynamic face stimuli

Participants performed a one-back repetition detection task during scanning.

Full details of stimuli, procedure, and analysis are provided in the paper.

## Related resources

Derived data, statistical maps, beta estimates, and full analysis code are available on OSF:

👉 https://osf.io/q2v68

The OSF repository includes:
- Single-trial beta estimates
- ROI masks
- Decoding analysis scripts (Python)
- Reproduction code for all figures in the paper

This OpenNeuro dataset contains only the **raw MRI data** required to reproduce those analyses.

## Data structure

The dataset follows the Brain Imaging Data Structure (BIDS) specification:

- `sub-*/ses-*/anat/` – T1-weighted anatomical images
- `sub-*/ses-*/func/` – Functional runs (main task and localizer)
- `events.tsv` files – Trial-level timing and condition labels
- `dataset_description.json` – Dataset metadata

## How to cite

If you use this dataset, please cite:

Hosseini, S. S., & Soto, F. A. (2025). *Invariant versus context-specific representation of face shape and motion in the face network.*

And include a link to this OpenNeuro dataset.

## License

This dataset is released under the **CC0 1.0 Public Domain Dedication**.
You are free to use, share, and adapt the data without restriction.

## Notes

- This dataset is intended for research and educational use.
- All participants provided informed consent for public data sharing.
- Faces in anatomical images have been removed via automated defacing.