Face Caricature

Face caricature research focuses on algorithmically generating exaggerated portraits that capture individual features while maintaining identity. Current efforts utilize generative adversarial networks (GANs), particularly StyleGAN, and explore techniques like latent space manipulation and cross-attention mechanisms to achieve high-quality, realistic caricatures from single images or sketches, even accommodating pose variations and occlusions. This work aims to improve the realism and controllability of automated caricature generation, potentially impacting fields like entertainment, art, and even forensic applications through enhanced image manipulation and synthesis capabilities.