CeLuigi / voxsrc_2019

VoxSRC Challenge

VoxSRC Challenge

This repository contains the script to compute Equal Error Rates (EER)

1. Download the list of validation pairs.

sh download_list.sh

2. The output should be formatted as below. See data/veri_test_output.txt for examples.

SCORE FILE1 FILE2

3. Run evaluate.py to compute the EER.

python evaluate.py --ground_truth=data/veri_test.txt --prediction=data/veri_test_output.txt --positive=0

Follow this link for more information about the challenge.