Oriol Ramos Terrades

TDIR reveals that you can effectively disentangle temporal and categorical information in image embeddings, enabling unprecedented flexibility in historical image retrieval.

Historical document retrieval can be both accurate and flexible, thanks to a novel agentic system that outperforms traditional methods in noisy environments.

Modeling the temporal dynamics of holograms can significantly enhance the security of identity verification systems against sophisticated attacks.

Formal verification meets deep learning in a new image retrieval framework that not only finds matching images, but also pinpoints exactly which query constraints are satisfied or unmet.