Roman Bachmann

Roman Bachmann is a Researcher at the École Polytechnique Fédérale de Lausanne (EPFL), affiliated with the Visual Intelligence and Learning Lab (VILAB) within the School of Computer and Communication Sciences (IC). He holds a doctoral status as an Assistant-doctorant in the Programme doctoral en informatique et communications. His research focuses on advanced computer vision, machine learning, and AI-driven multimodal systems, with contributions to generative models, vision-language integration, and 3D scanning technologies.

Key research interests include visual personalization (ViPer), multimodal learning (4M series), and scalable vision pipelines (Omnidata). His work bridges theoretical advancements with practical applications in embodied AI, robotics, and creative technologies. He is actively involved in the VILAB, contributing to projects that address challenges in cross-modal understanding and adaptive systems.

Roman’s publications reflect a strong emphasis on innovation in AI, with recent work exploring flexible tokenization (Flextok) and task-agnostic vision models (4M-21). His research has implications for robotics, generative AI, and data-driven decision-making in complex environments.