Model evaluation harness for vision model evaluation on the Fair human-centric image dataset for ethical AI benchmarking (FHIBE) dataset developed by the Sony AI Ethics Team. - View it on GitHub