Florian Strub

Papers

Total Citations

H-Index

About

Florian Strub is a leading researcher in artificial intelligence, specializing in multimodal learning, reinforcement learning, and grounded language understanding. His work bridges the gap between perception and interaction, enabling AI agents to learn from vision, audio, semantics, and physical environments. Strub is best known for introducing **HoME: a Household Multimodal Environment** (2017, over 80 citations), a pioneering simulation platform that integrates over 45,000 realistic 3D house layouts from the SUNCG dataset. This environment allows agents to learn through vision, sound, physics, and object interaction, setting a new standard for embodied AI research. His contributions have significantly advanced the field of multimodal representation learning, demonstrating how agents can acquire language and reasoning skills by exploring complex, interactive spaces. Strub’s work has been widely cited and continues to influence research in robotics, virtual agents, and human-AI interaction, making him a key figure in the development of more intelligent, context-aware artificial systems.