Comparative Analysis of Generative Networks for Synthesis of Images in Different Visual Domains

• Grant number: 25/01757-5
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: June 01, 2025
• End date: December 31, 2025
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Danillo Roberto Pereira
• Grantee: Marco Vinicius de Melo Faria
• Host Institution: Faculdade de Ciências e Tecnologia (FCT). Universidade Estadual Paulista (UNESP). Campus de Presidente Prudente. Presidente Prudente , SP, Brazil

Abstract

Artificial Intelligence (AI) has revolutionized various fields, especially synthetic data generation. Within this context, generative networks have emerged as powerful tools capable of creating realistic images based on existing datasets. The three main generative networks today are: (i) Generative Adversarial Networks (GANs), (ii) Conditional Generative Adversarial Networks (CGANs), and (iii) Variational Autoencoders (VAE). Each of these networks employs different approaches to image generation, impacting the quality and versatility of the results. These three generative networks aim to generate synthetic data as close to reality as possible using distinct approaches. This Scientific Initiation project aims to compare these three networks on different datasets and conduct a comparative analysis to evaluate their performance, accuracy, and versatility. (AU)