A Comparative Study of Image-to-Image Translation Using GANs for Synthetic Child Race Data

TL;DR

This paper explores using GAN-based image-to-image translation methods to synthesize ethnically diverse children's face data, addressing data scarcity and improving face recognition across different ethnicities.

Contribution

It compares three GAN-based image-to-image translation methods for synthesizing diverse child face data, demonstrating their feasibility in enhancing ethnic diversity in datasets.

Findings

Synthetic data successfully generated with broader ethnic diversity.

Image-to-image translation methods are effective for data augmentation.

Feasibility demonstrated on synthetic datasets.

Abstract

The lack of ethnic diversity in data has been a limiting factor of face recognition techniques in the literature. This is particularly the case for children where data samples are scarce and presents a challenge when seeking to adapt machine vision algorithms that are trained on adult data to work on children. This work proposes the utilization of image-to-image transformation to synthesize data of different races and thus adjust the ethnicity of children's face data. We consider ethnicity as a style and compare three different Image-to-Image neural network based methods, specifically pix2pix, CycleGAN, and CUT networks to implement Caucasian child data and Asian child data conversion. Experimental validation results on synthetic data demonstrate the feasibility of using image-to-image transformation methods to generate various synthetic child data samples with broader ethnic diversity.