Mengling Feng

Semantic features derived from Visual Question Answering can dramatically enhance the robustness of cancer prognosis models across diverse clinical settings.

Overconfident predictions plague mental health prediction models, but this new framework leverages evidential learning to provide more trustworthy uncertainty estimates and human-understandable reasoning signals.

EEG foundation models struggle with limited data because of structural mismatch, but SCOPE solves this by creating structured prototypes that guide adaptation, leading to better generalization.