Professionally Diverse: AI-Generated Faces for Targeted Advertising

We explore the application of neural networks in generating realistic and diverse faces for the advertisement industry. The study involves a thorough literature review of generative models up to StyleGAN2, and investigating its potential impact on the creation of compelling visuals for advertising campaigns. We develop a new framework for generating images for targeted advertising with automatic face detection. We propose a method for collecting images for creating massive training datasets. We check some variants of generative models to arrive at using CNNs with additional style modules and the Frechet inception distance.