معرفی
Federico Becattini is a Tenure-Track Assistant Professor at the Department of Information Engineering and Mathematics (DIISM), University of Siena, Italy. He is an active member of the Siena Artificial Intelligence Lab (SAILab), where he contributes to cutting-edge research in computer vision, deep learning, and artificial intelligence. His work spans multiple interdisciplinary domains, including autonomous driving, human behavior understanding, cultural heritage, neuromorphic vision, and fashion recommendation.
His research interests center on memory-based neural networks, which he has applied in numerous publications at top-tier venues such as CVPR, ECCV, IEEE TPAMI, and ACM TOMM. He has also delivered tutorials on this topic at international conferences including ICIAP 2022 and ACM MM 2022, and taught a Ph.D. course at the University of Florence. His recent work is aligned with the Collectionless AI paradigm, which emphasizes continual learning and interaction with dynamic environments.
The recent publications highlight a strong trend in human-centric AI, focusing on understanding people through multimodal analysis of face, body, and clothing, as well as generating 3D virtual avatars. There is also a clear emphasis on memory-augmented architectures for temporal reasoning, explainability, and adaptive learning. His editorial role as Associate Editor of the International Journal of Multimedia Information Retrieval further underscores his standing in the research community.
- Associate Editor, International Journal of Multimedia Information Retrieval (IJMIR)
- Organizer, Workshop on Facial and Body Expressions (ICPR2020)
- Co-organizer, T-CAP Workshop (ICIAP2021, ICPR2022)
- Co-organizer, MCFR Workshop (ACM MM 2022)
- Co-organizer, WCPA Workshop and Challenge (ECCV 2022)
Federico Becattini actively advises students and researchers within SAILab, particularly in the context of Ph.D. theses and research projects related to Collectionless AI and memory-based models. While specific grants are not mentioned, his extensive publication record and leadership in workshops and editorial roles suggest involvement in funded research initiatives. He collaborates with both academic and international research communities, serving as a reviewer for top-tier conferences and journals.
He is a core member of the SAILab research group, which is pioneering the Collectionless AI initiative—a framework for continual learning over time, interacting with humans and agents without relying on pre-built static datasets. This lab serves as a hub for innovation in adaptive and sustainable AI systems.
