The main folder contains all files relavant to the analysis of this paper. The `data` file contains publicly available data used in the manuscript. Individuals wishing to obtain data not available for public release shall contact the corresponding author of this paper. The `fairface_evaluation.csv` contains the manual versus computational character race codings. The `analysis_osf.ipynb` jupyter notebook and `models_R_share.Rmd` markdown details the various analyses of the paper.
The `notebooks_instructional` folder contains tutorials for the character recognition, facial extraction, and feature construction pipelines that were implemented in this paper. The `configurations_instructional` folder contains all relevant materials associated with these tutorials. Due to file size limits, we have not uploaded the yolo pretrained model weights mentioned in our facial extraction pipeline. You may download these weights directly from the relevant github repository mentioned in the notebook.