Data loading package and documentation for the LAION-fMRI dataset
LAION-fMRI is a deeply sampled 7T fMRI dataset of brain responses to natural images, collected and curated by ViCCo-Group. This repository contains the Python package and documentation for dataset access, structure, specifications, and usage guidelines.
Full documentation is available at https://laion-fmri.hebartlab.com/ and includes:
- Dataset specifications and structure
- Access and usage guidelines
- Technical details and metadata
- Contributing guidelines
Until the PyPI package is published, install directly from GitHub:
python -m pip install "laion-fmri @ git+https://github.com/ViCCo-Group/LAION-fMRI.git@main"The package is imported as laion_fmri; the command-line tool is
laion-fmri.
This repository uses Sphinx for documentation. To set up a development environment:
# Create a conda environment from the development environment file
conda env create -f environment.yml
conda activate laion-fmri-dev
# Build the documentation
cd docs
make htmlThe built documentation will be available in docs/build/html/.
The package code is licensed under BSD-3-Clause. The fMRI data and public derivatives are released under CC0 1.0. Raw stimulus images require the LAION-fMRI Data Use Agreement.
Contributions are welcome! Please see the Contributing Guidelines in the documentation for details.
- Issues: GitHub Issues
- Discussions: GitHub Discussions