GENERATION AND DETECTION OF ARTIFICIAL FACIAL IMAGES: AN EXPLORATION

This project investigates the surge in AI-generated images, particularly the realistic portraits of human faces. While offering creative and entertainment possibilities, it raises concerns about trust and authenticity in digital content due to the potential misuse of these AI-generated contents which can be used for creating deceptive identities. The dual-edge nature that this technology provides, are the critical tools that can distinguish artificial faces from genuine ones. Our objective is a comprehensive exploration of techniques for both generating and detecting artificial facial images, evaluating accuracy score for each models that we list below. Generation: DALL-E2, Mid Journey, Stable Diffusion Detection: Resnet-50, VGG-16, Inception, Xception, AlexNet. The outcome of this research will help the researchers and students all around the world with their ongoing discussions on responsible AI development.