Using Augmented Face Images to Improve Facial Recognition Tasks

TL;DR

This paper introduces a framework that employs GAN-generated augmented face images to enhance the training process, especially for underrepresented attributes, thereby improving facial recognition accuracy.

Contribution

The novel contribution is the use of GAN-augmented images targeting specific underrepresented attributes to boost facial recognition performance.

Findings

Improved recognition accuracy for underrepresented attributes.

Effective augmentation method using GANs.

Enhanced training dataset diversity.

Abstract

We present a framework that uses GAN-augmented images to complement certain specific attributes, usually underrepresented, for machine learning model training. This allows us to improve inference quality over those attributes for the facial recognition tasks.