DynProto dynamically captures coarse OOD patterns confused with each ID class and refines them into prototypes during testing to enable similarity-based OOD detection that outperforms prior methods on ImageNet benchmarks.
Fa: Forced prompt learning of vision-language models for out-of-distribution detection
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
DynProto: Dynamic Prototype Evolution for Out-of-Distribution Detection
DynProto dynamically captures coarse OOD patterns confused with each ID class and refines them into prototypes during testing to enable similarity-based OOD detection that outperforms prior methods on ImageNet benchmarks.