Long Bio
Marco Mistretta, born in Arezzo in 1999, is a PhD student and research fellow at the Media Integration and Communication Center (MICC) at the University of Florence, under the supervision of Prof. Andrew D. Bagdanov and Prof. Marco Bertini. His research focuses on Multimodal Vision-Language Models, Incremental Learning, and Prompt Learning, with an emphasis on bridging the gap between cutting-edge AI research and practical applications. Marco’s work explores novel methodologies for improving the adaptability and efficiency of AI systems in real-world scenarios, particularly in zero-shot and few-shot learning contexts.
Marco earned his Bachelor’s degree in Computer Engineering from the University of Florence in 2021, where he presented his thesis titled “Scarlatti-Gen: AI-Driven Sonata Generation Using Weighted Graphs and CNNs,” under the supervision of Prof. Paolo Frasconi and Prof. Simone Conforti, achieving a final grade of 105/110. He continued his academic journey at the same university, completing his Master’s degree in Artificial Intelligence in 2023 with a thesis titled “RE-Tune: Incremental Fine-Tuning of Biomedical Vision-Language Models,” under the supervision of Prof. Andrew D. Bagdanov and Prof. Marco Bertini, graduating with the highest honors (110/110L).
Marco has made significant contributions to the field of Artificial Intelligence and Computer Vision, publishing as the first author in top-tier venues such as ECCV and NeurIPS. His recent works include KDPL, a novel approach to improving zero-shot generalization of learned prompts via unsupervised knowledge distillation, and RE-Tune, an incremental fine-tuning strategy for biomedical vision-language models. These projects reflect his passion for addressing real-world challenges like privacy-preserving healthcare AI and efficient model adaptation without labeled data.