Behzad Bozorgtabar

Behzad Bozorgtabar is a Lecturer at École Polytechnique Fédérale de Lausanne (EPFL) and a Senior Scientist at the Lausanne University Hospital (CHUV). He leads the Medical Imaging Group at EPFL's Signal Processing Laboratory (LTS5), with affiliations to the European Lab for Learning & Intelligent Systems (ELLIS) and EPFL’s ELLIS Unit.

His research focuses on self-supervised learning, test-time adaptation, and multimodal foundation models, aiming to create robust AI systems that adapt to changing environments. His work addresses domain shift, anomaly detection, and ethical AI deployment in medical imaging and computational pathology.

His recent publications highlight trends in test-time adaptation (e.g., ReservoirTTA, OptTTA), self-supervised learning (CrIBo, CrOC), and medical imaging applications (histopathology, chest X-rays, colorectal cancer detection).

Scientific Awards:

• IEEE TMI Distinguished Reviewer Award
• First Runner-up - AIMIA Workshop, ECCV
• Elected Member - European Lab for Learning & Intelligent Systems (ELLIS)
• IBM Research Division Image Award
• IBM Patent Award
• Imagine Cup, Australian Finals, Microsoft

Advising & Funding: He has mentored PhD students (Tomar, Vray, Stegmüller) and secured grants from the Personalized Health and Related Technologies (PHRT), Swiss Cancer League, and Discovery Translation Fund (DTF 2.0). His service includes Area Chair roles for CVPR 2025, WACV 2026, and AAAI 2026.