Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313194958> ?p ?o ?g. }
- W4313194958 endingPage "368" @default.
- W4313194958 startingPage "351" @default.
- W4313194958 abstract "Existing prototypical-based models address the black-box nature of deep learning. However, they are sub-optimal as they often assume separate prototypes for each class, require multi-step optimization, make decisions based on prototype absence (so-called negative reasoning process), and derive vague prototypes. To address those shortcomings, we introduce ProtoPool, an interpretable prototype-based model with positive reasoning and three main novelties. Firstly, we reuse prototypes in classes, which significantly decreases their number. Secondly, we allow automatic, fully differentiable assignment of prototypes to classes, which substantially simplifies the training process. Finally, we propose a new focal similarity function that contrasts the prototype from the background and consequently concentrates on more salient visual features. We show that ProtoPool obtains state-of-the-art accuracy on the CUB-200-2011 and the Stanford Cars datasets, substantially reducing the number of prototypes. We provide a theoretical analysis of the method and a user study to show that our prototypes capture more salient features than those obtained with competitive methods. We made the code available at https://github.com/gmum/ProtoPool ." @default.
- W4313194958 created "2023-01-06" @default.
- W4313194958 creator A5001765250 @default.
- W4313194958 creator A5007063171 @default.
- W4313194958 creator A5024637012 @default.
- W4313194958 creator A5026834372 @default.
- W4313194958 creator A5052791052 @default.
- W4313194958 creator A5058095594 @default.
- W4313194958 date "2022-01-01" @default.
- W4313194958 modified "2023-10-14" @default.
- W4313194958 title "Interpretable Image Classification with Differentiable Prototypes Assignment" @default.
- W4313194958 cites W1964120672 @default.
- W4313194958 cites W2045830397 @default.
- W4313194958 cites W2108598243 @default.
- W4313194958 cites W2138011018 @default.
- W4313194958 cites W2194775991 @default.
- W4313194958 cites W2282821441 @default.
- W4313194958 cites W2765813195 @default.
- W4313194958 cites W2773003563 @default.
- W4313194958 cites W2797977484 @default.
- W4313194958 cites W2895739182 @default.
- W4313194958 cites W2942678267 @default.
- W4313194958 cites W2945295328 @default.
- W4313194958 cites W2945976633 @default.
- W4313194958 cites W2954471638 @default.
- W4313194958 cites W2962858109 @default.
- W4313194958 cites W2963407932 @default.
- W4313194958 cites W2963446712 @default.
- W4313194958 cites W2983256121 @default.
- W4313194958 cites W2988157455 @default.
- W4313194958 cites W3004725381 @default.
- W4313194958 cites W3012380364 @default.
- W4313194958 cites W3034873438 @default.
- W4313194958 cites W3035145964 @default.
- W4313194958 cites W3035393070 @default.
- W4313194958 cites W3037211332 @default.
- W4313194958 cites W3037346646 @default.
- W4313194958 cites W3101349400 @default.
- W4313194958 cites W3101609372 @default.
- W4313194958 cites W3135579708 @default.
- W4313194958 cites W3167738759 @default.
- W4313194958 cites W3172917901 @default.
- W4313194958 cites W3175314453 @default.
- W4313194958 cites W3177934633 @default.
- W4313194958 cites W3191974319 @default.
- W4313194958 cites W3202988816 @default.
- W4313194958 cites W3209259671 @default.
- W4313194958 cites W3215430231 @default.
- W4313194958 cites W4200254023 @default.
- W4313194958 cites W4214561053 @default.
- W4313194958 cites W4241027858 @default.
- W4313194958 doi "https://doi.org/10.1007/978-3-031-19775-8_21" @default.
- W4313194958 hasPublicationYear "2022" @default.
- W4313194958 type Work @default.
- W4313194958 citedByCount "5" @default.
- W4313194958 countsByYear W43131949582023 @default.
- W4313194958 crossrefType "book-chapter" @default.
- W4313194958 hasAuthorship W4313194958A5001765250 @default.
- W4313194958 hasAuthorship W4313194958A5007063171 @default.
- W4313194958 hasAuthorship W4313194958A5024637012 @default.
- W4313194958 hasAuthorship W4313194958A5026834372 @default.
- W4313194958 hasAuthorship W4313194958A5052791052 @default.
- W4313194958 hasAuthorship W4313194958A5058095594 @default.
- W4313194958 hasBestOaLocation W43131949582 @default.
- W4313194958 hasConcept C103278499 @default.
- W4313194958 hasConcept C115961682 @default.
- W4313194958 hasConcept C119857082 @default.
- W4313194958 hasConcept C124101348 @default.
- W4313194958 hasConcept C134306372 @default.
- W4313194958 hasConcept C14036430 @default.
- W4313194958 hasConcept C154945302 @default.
- W4313194958 hasConcept C177264268 @default.
- W4313194958 hasConcept C18903297 @default.
- W4313194958 hasConcept C199360897 @default.
- W4313194958 hasConcept C202615002 @default.
- W4313194958 hasConcept C206588197 @default.
- W4313194958 hasConcept C2776760102 @default.
- W4313194958 hasConcept C2777212361 @default.
- W4313194958 hasConcept C2780719617 @default.
- W4313194958 hasConcept C33923547 @default.
- W4313194958 hasConcept C41008148 @default.
- W4313194958 hasConcept C78458016 @default.
- W4313194958 hasConcept C86803240 @default.
- W4313194958 hasConcept C98045186 @default.
- W4313194958 hasConceptScore W4313194958C103278499 @default.
- W4313194958 hasConceptScore W4313194958C115961682 @default.
- W4313194958 hasConceptScore W4313194958C119857082 @default.
- W4313194958 hasConceptScore W4313194958C124101348 @default.
- W4313194958 hasConceptScore W4313194958C134306372 @default.
- W4313194958 hasConceptScore W4313194958C14036430 @default.
- W4313194958 hasConceptScore W4313194958C154945302 @default.
- W4313194958 hasConceptScore W4313194958C177264268 @default.
- W4313194958 hasConceptScore W4313194958C18903297 @default.
- W4313194958 hasConceptScore W4313194958C199360897 @default.
- W4313194958 hasConceptScore W4313194958C202615002 @default.
- W4313194958 hasConceptScore W4313194958C206588197 @default.
- W4313194958 hasConceptScore W4313194958C2776760102 @default.
- W4313194958 hasConceptScore W4313194958C2777212361 @default.