Overview

This repository computes False Acceptance Rate (FAR), False Rejection Rate (FRR) and Equal Error Rate (EER) given similarity scores of a biometric system and the actual label of the sample (genuine or imposter). It also plots some visualizations such as histogram of genuine (blue) and imposter (red) distributions with a given bin width, EER along with FAR and FRR, and ROC curve.

Installation

1. Git clone this repo: git clone https://github.com/nathanlem1/Compute_EER.git
2. Install dependencies by pip install -r requirements.txt to have the same environment configuration as the one we used.

Run

$ python compute_eer.py