Why Feature Magnitude Deceives OOD Detectors: An Angular Separation Perspective
Reveals why feature magnitude misleads OOD detectors and proposes an angular separation perspective for robust out-of-distribution detection.
You can also find my articles on my Google Scholar profile.
Reveals why feature magnitude misleads OOD detectors and proposes an angular separation perspective for robust out-of-distribution detection.
Proposes PTTA, a method that purifies malicious samples to enable robust test-time model adaptation.
Proposes decoupled entropy minimization for improved test-time adaptation and domain generalization.
Introduces a frequency-aware approach for domain generalization by leveraging spectral domain information.