Matches in SemOpenAlex for { <https://semopenalex.org/work/W3201670937> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W3201670937 abstract "Deep CNNs, though have achieved the state of the art performance in image classification tasks, remain a black-box to a human using them. There is a growing interest in explaining the working of these deep models to improve their trustworthiness. In this paper, we introduce a Posthoc Architecture-agnostic Concept Extractor (PACE) that automatically extracts smaller sub-regions of the image called concepts relevant to the black-box prediction. PACE tightly integrates the faithfulness of the explanatory framework to the black-box model. To the best of our knowledge, this is the first work that extracts class-specific discriminative concepts in a posthoc manner automatically. The PACE framework is used to generate explanations for two different CNN architectures trained for classifying the AWA2 and Imagenet- Birds datasets. Extensive human subject experiments are conducted to validate the human interpretability and consistency of the explanations extracted by PACE. The results from these experiments suggest that over 72% of the concepts extracted by PACE are human interpretable." @default.
- W3201670937 created "2021-10-11" @default.
- W3201670937 creator A5000823166 @default.
- W3201670937 creator A5033210021 @default.
- W3201670937 creator A5054278782 @default.
- W3201670937 date "2021-07-18" @default.
- W3201670937 modified "2023-10-13" @default.
- W3201670937 title "PACE: Posthoc Architecture-Agnostic Concept Extractor for Explaining CNNs" @default.
- W3201670937 cites W2117539524 @default.
- W3201670937 cites W2194775991 @default.
- W3201670937 cites W2282821441 @default.
- W3201670937 cites W2295107390 @default.
- W3201670937 cites W2765793020 @default.
- W3201670937 cites W2776207810 @default.
- W3201670937 cites W2788403449 @default.
- W3201670937 cites W2954471638 @default.
- W3201670937 cites W2959587146 @default.
- W3201670937 cites W2962858109 @default.
- W3201670937 cites W2963499153 @default.
- W3201670937 cites W2988157455 @default.
- W3201670937 cites W3010219363 @default.
- W3201670937 cites W3035253074 @default.
- W3201670937 cites W3035520720 @default.
- W3201670937 cites W3099206234 @default.
- W3201670937 cites W3101609372 @default.
- W3201670937 doi "https://doi.org/10.1109/ijcnn52387.2021.9534369" @default.
- W3201670937 hasPublicationYear "2021" @default.
- W3201670937 type Work @default.
- W3201670937 sameAs 3201670937 @default.
- W3201670937 citedByCount "6" @default.
- W3201670937 countsByYear W32016709372022 @default.
- W3201670937 countsByYear W32016709372023 @default.
- W3201670937 crossrefType "proceedings-article" @default.
- W3201670937 hasAuthorship W3201670937A5000823166 @default.
- W3201670937 hasAuthorship W3201670937A5033210021 @default.
- W3201670937 hasAuthorship W3201670937A5054278782 @default.
- W3201670937 hasBestOaLocation W32016709372 @default.
- W3201670937 hasConcept C115961682 @default.
- W3201670937 hasConcept C119857082 @default.
- W3201670937 hasConcept C123657996 @default.
- W3201670937 hasConcept C13280743 @default.
- W3201670937 hasConcept C142362112 @default.
- W3201670937 hasConcept C153180895 @default.
- W3201670937 hasConcept C153349607 @default.
- W3201670937 hasConcept C154945302 @default.
- W3201670937 hasConcept C205649164 @default.
- W3201670937 hasConcept C2776436953 @default.
- W3201670937 hasConcept C2777526511 @default.
- W3201670937 hasConcept C2781067378 @default.
- W3201670937 hasConcept C41008148 @default.
- W3201670937 hasConcept C75294576 @default.
- W3201670937 hasConcept C97931131 @default.
- W3201670937 hasConceptScore W3201670937C115961682 @default.
- W3201670937 hasConceptScore W3201670937C119857082 @default.
- W3201670937 hasConceptScore W3201670937C123657996 @default.
- W3201670937 hasConceptScore W3201670937C13280743 @default.
- W3201670937 hasConceptScore W3201670937C142362112 @default.
- W3201670937 hasConceptScore W3201670937C153180895 @default.
- W3201670937 hasConceptScore W3201670937C153349607 @default.
- W3201670937 hasConceptScore W3201670937C154945302 @default.
- W3201670937 hasConceptScore W3201670937C205649164 @default.
- W3201670937 hasConceptScore W3201670937C2776436953 @default.
- W3201670937 hasConceptScore W3201670937C2777526511 @default.
- W3201670937 hasConceptScore W3201670937C2781067378 @default.
- W3201670937 hasConceptScore W3201670937C41008148 @default.
- W3201670937 hasConceptScore W3201670937C75294576 @default.
- W3201670937 hasConceptScore W3201670937C97931131 @default.
- W3201670937 hasLocation W32016709371 @default.
- W3201670937 hasLocation W32016709372 @default.
- W3201670937 hasOpenAccess W3201670937 @default.
- W3201670937 hasPrimaryLocation W32016709371 @default.
- W3201670937 hasRelatedWork W2761785940 @default.
- W3201670937 hasRelatedWork W2806259446 @default.
- W3201670937 hasRelatedWork W2905433371 @default.
- W3201670937 hasRelatedWork W2971377935 @default.
- W3201670937 hasRelatedWork W4310278675 @default.
- W3201670937 hasRelatedWork W4311431240 @default.
- W3201670937 hasRelatedWork W4312407344 @default.
- W3201670937 hasRelatedWork W4361193272 @default.
- W3201670937 hasRelatedWork W4384115502 @default.
- W3201670937 hasRelatedWork W2963326959 @default.
- W3201670937 isParatext "false" @default.
- W3201670937 isRetracted "false" @default.
- W3201670937 magId "3201670937" @default.
- W3201670937 workType "article" @default.