Matches in SemOpenAlex for { <https://semopenalex.org/work/W3136015650> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W3136015650 abstract "Neural networks trained to classify images do so by identifying features that allow them to distinguish between classes. These sets of features are either causal or context dependent. Grad-CAM is a popular method of visualizing both sets of features. In this paper, we formalize this feature divide and provide a methodology to extract causal features from Grad-CAM. We do so by defining context features as those features that allow contrast between predicted class and any contrast class. We then apply a set theoretic approach to separate causal from contrast features for COVID-19 CT scans. We show that on average, the image regions with the proposed causal features require 15% less bits when encoded using Huffman encoding, compared to Grad-CAM, for an average increase of 3% classification accuracy, over Grad-CAM. Moreover, we validate the transfer-ability of causal features between networks and comment on the non-human interpretable causal nature of current networks." @default.
- W3136015650 created "2021-03-29" @default.
- W3136015650 creator A5006145139 @default.
- W3136015650 creator A5010958338 @default.
- W3136015650 date "2021-03-23" @default.
- W3136015650 modified "2023-09-27" @default.
- W3136015650 title "Extracting Causal Visual Features for Limited label Classification" @default.
- W3136015650 cites W1985907861 @default.
- W3136015650 cites W2060108852 @default.
- W3136015650 cites W2081625202 @default.
- W3136015650 cites W2117539524 @default.
- W3136015650 cites W2123045220 @default.
- W3136015650 cites W2159564241 @default.
- W3136015650 cites W2194775991 @default.
- W3136015650 cites W2765793020 @default.
- W3136015650 cites W2892074823 @default.
- W3136015650 cites W2950772592 @default.
- W3136015650 cites W2963446712 @default.
- W3136015650 cites W2963528347 @default.
- W3136015650 cites W2963941635 @default.
- W3136015650 cites W2970652577 @default.
- W3136015650 cites W3014337038 @default.
- W3136015650 cites W3017243633 @default.
- W3136015650 cites W3090434739 @default.
- W3136015650 cites W3091150073 @default.
- W3136015650 cites W3091757450 @default.
- W3136015650 cites W3098219814 @default.
- W3136015650 cites W3110039477 @default.
- W3136015650 hasPublicationYear "2021" @default.
- W3136015650 type Work @default.
- W3136015650 sameAs 3136015650 @default.
- W3136015650 citedByCount "0" @default.
- W3136015650 crossrefType "posted-content" @default.
- W3136015650 hasAuthorship W3136015650A5006145139 @default.
- W3136015650 hasAuthorship W3136015650A5010958338 @default.
- W3136015650 hasConcept C119857082 @default.
- W3136015650 hasConcept C134306372 @default.
- W3136015650 hasConcept C138885662 @default.
- W3136015650 hasConcept C151730666 @default.
- W3136015650 hasConcept C153180895 @default.
- W3136015650 hasConcept C154945302 @default.
- W3136015650 hasConcept C177148314 @default.
- W3136015650 hasConcept C177264268 @default.
- W3136015650 hasConcept C199360897 @default.
- W3136015650 hasConcept C2776401178 @default.
- W3136015650 hasConcept C2776502983 @default.
- W3136015650 hasConcept C2777212361 @default.
- W3136015650 hasConcept C2779343474 @default.
- W3136015650 hasConcept C33923547 @default.
- W3136015650 hasConcept C41008148 @default.
- W3136015650 hasConcept C41895202 @default.
- W3136015650 hasConcept C50644808 @default.
- W3136015650 hasConcept C86803240 @default.
- W3136015650 hasConceptScore W3136015650C119857082 @default.
- W3136015650 hasConceptScore W3136015650C134306372 @default.
- W3136015650 hasConceptScore W3136015650C138885662 @default.
- W3136015650 hasConceptScore W3136015650C151730666 @default.
- W3136015650 hasConceptScore W3136015650C153180895 @default.
- W3136015650 hasConceptScore W3136015650C154945302 @default.
- W3136015650 hasConceptScore W3136015650C177148314 @default.
- W3136015650 hasConceptScore W3136015650C177264268 @default.
- W3136015650 hasConceptScore W3136015650C199360897 @default.
- W3136015650 hasConceptScore W3136015650C2776401178 @default.
- W3136015650 hasConceptScore W3136015650C2776502983 @default.
- W3136015650 hasConceptScore W3136015650C2777212361 @default.
- W3136015650 hasConceptScore W3136015650C2779343474 @default.
- W3136015650 hasConceptScore W3136015650C33923547 @default.
- W3136015650 hasConceptScore W3136015650C41008148 @default.
- W3136015650 hasConceptScore W3136015650C41895202 @default.
- W3136015650 hasConceptScore W3136015650C50644808 @default.
- W3136015650 hasConceptScore W3136015650C86803240 @default.
- W3136015650 hasLocation W31360156501 @default.
- W3136015650 hasOpenAccess W3136015650 @default.
- W3136015650 hasPrimaryLocation W31360156501 @default.
- W3136015650 hasRelatedWork W1537370018 @default.
- W3136015650 hasRelatedWork W1777638167 @default.
- W3136015650 hasRelatedWork W1990980378 @default.
- W3136015650 hasRelatedWork W2081094030 @default.
- W3136015650 hasRelatedWork W2095166651 @default.
- W3136015650 hasRelatedWork W2115638326 @default.
- W3136015650 hasRelatedWork W2140446226 @default.
- W3136015650 hasRelatedWork W2166430258 @default.
- W3136015650 hasRelatedWork W2507794850 @default.
- W3136015650 hasRelatedWork W2572038443 @default.
- W3136015650 hasRelatedWork W2614796172 @default.
- W3136015650 hasRelatedWork W2768204113 @default.
- W3136015650 hasRelatedWork W2896145961 @default.
- W3136015650 hasRelatedWork W2947469701 @default.
- W3136015650 hasRelatedWork W2963258075 @default.
- W3136015650 hasRelatedWork W2969139878 @default.
- W3136015650 hasRelatedWork W3193999289 @default.
- W3136015650 hasRelatedWork W37721291 @default.
- W3136015650 hasRelatedWork W79324960 @default.
- W3136015650 hasRelatedWork W1697230285 @default.
- W3136015650 isParatext "false" @default.
- W3136015650 isRetracted "false" @default.
- W3136015650 magId "3136015650" @default.
- W3136015650 workType "article" @default.