Arc2Avatar: Generating Expressive 3D Avatars from a Single Image via ID Guidance, CVPR 2025; AnimateMe: 4D Facial Expressions via Diffusion Models, ECCV 2024.
Research Experience
Conducting PhD research at the Department of Computing, Imperial College London, focusing on 3D facial reconstruction and generating expressive 3D avatars.
Education
PhD Candidate, Imperial College London, supervised by Professor Stefanos Zafeiriou; Diploma (BSc+MEng) in Computer Science, University of Patras, advised by Professor Emmanouil Z. Psarakis for his thesis.
Background
Research interests: Machine Learning, Computer Vision, and Graphics, with a particular focus on Digital Humans. Works on methods for 3D facial reconstruction from limited-view setups (as little as a single image) to create realistic avatars using 3D representations (3D Gaussian Splats) and generative models (Diffusion Models). Also focuses on making these avatars controllable to convey expressions, speech, and animation.