Hi! I am Fulvio Sanguigni! I am an Italian National Phd student, part of AImageLab at University of Modena and Reggio Emilia, supervised by Marcella Cornia and Lorenzo Baraldi My research direction is focused both towards leveraging pre-trained, multimodal models for new image editing applications and addressing some key, structural problems of these architectures. The targets are:

• Fashion Domain. Application to virtual try-on and garment editing, both clothed humans and in-shop garments.
• Generative AI robustness and alignment: Investigating architectural leaks in generative models, adaptability across domains and sampling efficiency, theoretical foundations.
• Multimodal Large Language Models: 1) provide context and reasoning to a vision generative model 2) improve multimodal capabilities of current LLMs

I have further interests combined with generative models, expecially regarding human motion analysis, feature extraction (like SLIME), and remote sensing (see also paragraph below)

I have a background in Artificial Intelligence and Robotics from Sapienza University of Rome. I have visited Φ-Lab at European Space Agency in 2022-2023 as a thesist. I graduated with a thesis on Diffusion Models for Earth Observation Data, supervised by Irene Amerini (Sapienza) and co-supervised by Bertrand Le Saux (Φ-Lab). Paper and Code

In the meanwhile, I delve into sports firstly for passion, secondly for the remarkable benefits of living a student-athlete life and finally for sake of competition. You can find more in my athlete page and in several blog posts.