Multiobjective evolutionary search of the latent space of Generative Adversarial Networks for human face generation

This article presents an explicit multiobjective evolutionary approach for synthetic human face image generation, exploring the latent space of generative adversarial networks. The approach considers the similarity to a target image and the race attribute. The evolutionary search explores the real-coded latent space of Style-GAN3 and applies DeepFace for similarity and race evaluation. Realistic images are generated, properly exploring the search space and the Pareto front of the problem. The generated images pose a challenge to the automatic detection system in DeepFace. Results are applicable to enhance the security of face recognition systems.