Impersonator++ (iPERCore) implements a Liquid Warping GAN with attention for human image synthesis. The framework enables three core capabilities: motion imitation (transferring pose from a reference to a source person), appearance transfer (applying source appearance to a target), and novel view synthesis (generating different viewing angles). It produces outputs at high resolution (512x512 or 1024x1024) using deep learning techniques and builds upon the original ICCV impersonator project with improved generalization.