Face Matching 1:1

Model by Trueface

This model can identify whether two images given as input are of the same person or not. The model uses a Deep Neural Network (DNN), which was trained end-to-end on large-scale face recognition datasets covering all ethnicities and genders.

  • Description

    Product Description

    PERFORMANCE METRICS:

    This model was evaluated against four datasets: LFW, agedb, cfp-fp, and cfp-ff.

    It achieves a true positive rate of 99.8% on LFW, 98.58% on agedb, 98.98% on cfp-fp, and 99.85% on cfp-ff.

    OVERVIEW:

    This model determines whether two images contain the same person or not.

    TRAINING:

    This model was trained on a dataset unique to Trueface that was compiled in-house from public and proprietary sources specifically to limit and combat biases present in publically available datasets alone.

    VALIDATION:

    This model ranks in the Top 10 for genuine template match on NIST 1:1 and was also tested on the Megaface, LFW, agedb, cfp-fp, and cfp-ff datasets and achieved state of the art performance.