Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171175978> ?p ?o ?g. }
- W3171175978 abstract "The class activation mapping, or CAM, has been the cornerstone of feature attribution methods for multiple vision tasks. Its simplicity and effectiveness have led to wide applications in the explanation of visual predictions and weakly-supervised localization tasks. However, CAM has its own shortcomings. The computation of attribution maps relies on ad-hoc calibration steps that are not part of the training computational graph, making it difficult for us to understand the real meaning of the attribution values. In this paper, we improve CAM by explicitly incorporating a latent variable encoding the location of the cue for recognition in the formulation, thereby subsuming the attribution map into the training computational graph. The resulting model, class activation latent mapping, or CALM, is trained with the expectation-maximization algorithm. Our experiments show that CALM identifies discriminative attributes for image classifiers more accurately than CAM and other visual attribution baselines. CALM also shows performance improvements over prior arts on the weakly-supervised object localization benchmarks. Our code is available at this https URL." @default.
- W3171175978 created "2021-06-22" @default.
- W3171175978 creator A5025851635 @default.
- W3171175978 creator A5040372929 @default.
- W3171175978 creator A5061578835 @default.
- W3171175978 creator A5078314427 @default.
- W3171175978 date "2021-06-15" @default.
- W3171175978 modified "2023-09-27" @default.
- W3171175978 title "Keep CALM and Improve Visual Feature Attribution" @default.
- W3171175978 cites W1533957530 @default.
- W3171175978 cites W1787224781 @default.
- W3171175978 cites W1903029394 @default.
- W3171175978 cites W1994488211 @default.
- W3171175978 cites W2021191744 @default.
- W3171175978 cites W2049633694 @default.
- W3171175978 cites W2081580037 @default.
- W3171175978 cites W2100470808 @default.
- W3171175978 cites W2117539524 @default.
- W3171175978 cites W2150165932 @default.
- W3171175978 cites W2195388612 @default.
- W3171175978 cites W2221898772 @default.
- W3171175978 cites W2240067561 @default.
- W3171175978 cites W2282821441 @default.
- W3171175978 cites W2295107390 @default.
- W3171175978 cites W2438305798 @default.
- W3171175978 cites W2503388974 @default.
- W3171175978 cites W2594633041 @default.
- W3171175978 cites W2605409611 @default.
- W3171175978 cites W2626639386 @default.
- W3171175978 cites W2657631929 @default.
- W3171175978 cites W2738853914 @default.
- W3171175978 cites W2765793020 @default.
- W3171175978 cites W2773497437 @default.
- W3171175978 cites W2788403449 @default.
- W3171175978 cites W2809136100 @default.
- W3171175978 cites W2809671526 @default.
- W3171175978 cites W2891612330 @default.
- W3171175978 cites W2895739182 @default.
- W3171175978 cites W2895976713 @default.
- W3171175978 cites W2945976633 @default.
- W3171175978 cites W2950557962 @default.
- W3171175978 cites W2952865063 @default.
- W3171175978 cites W2962772482 @default.
- W3171175978 cites W2962851944 @default.
- W3171175978 cites W2962858109 @default.
- W3171175978 cites W2962967409 @default.
- W3171175978 cites W2963045696 @default.
- W3171175978 cites W2963249138 @default.
- W3171175978 cites W2963382180 @default.
- W3171175978 cites W2963420481 @default.
- W3171175978 cites W2963483561 @default.
- W3171175978 cites W2963791246 @default.
- W3171175978 cites W2964274719 @default.
- W3171175978 cites W2970447476 @default.
- W3171175978 cites W2970852027 @default.
- W3171175978 cites W2971048680 @default.
- W3171175978 cites W2988157455 @default.
- W3171175978 cites W2992308087 @default.
- W3171175978 cites W3010219363 @default.
- W3171175978 cites W3035253074 @default.
- W3171175978 cites W3035602609 @default.
- W3171175978 cites W3035712793 @default.
- W3171175978 cites W3041935510 @default.
- W3171175978 cites W3080532707 @default.
- W3171175978 cites W3099579437 @default.
- W3171175978 cites W3101609372 @default.
- W3171175978 cites W3103376464 @default.
- W3171175978 cites W3108052345 @default.
- W3171175978 cites W3108643758 @default.
- W3171175978 cites W3116286104 @default.
- W3171175978 cites W3127579051 @default.
- W3171175978 cites W3140968660 @default.
- W3171175978 cites W14333344 @default.
- W3171175978 cites W2963326959 @default.
- W3171175978 hasPublicationYear "2021" @default.
- W3171175978 type Work @default.
- W3171175978 sameAs 3171175978 @default.
- W3171175978 citedByCount "1" @default.
- W3171175978 countsByYear W31711759782021 @default.
- W3171175978 crossrefType "posted-content" @default.
- W3171175978 hasAuthorship W3171175978A5025851635 @default.
- W3171175978 hasAuthorship W3171175978A5040372929 @default.
- W3171175978 hasAuthorship W3171175978A5061578835 @default.
- W3171175978 hasAuthorship W3171175978A5078314427 @default.
- W3171175978 hasConcept C119857082 @default.
- W3171175978 hasConcept C132525143 @default.
- W3171175978 hasConcept C138885662 @default.
- W3171175978 hasConcept C143299363 @default.
- W3171175978 hasConcept C153180895 @default.
- W3171175978 hasConcept C154945302 @default.
- W3171175978 hasConcept C15744967 @default.
- W3171175978 hasConcept C2776401178 @default.
- W3171175978 hasConcept C2777212361 @default.
- W3171175978 hasConcept C41008148 @default.
- W3171175978 hasConcept C41895202 @default.
- W3171175978 hasConcept C77805123 @default.
- W3171175978 hasConcept C80444323 @default.
- W3171175978 hasConcept C97931131 @default.
- W3171175978 hasConceptScore W3171175978C119857082 @default.
- W3171175978 hasConceptScore W3171175978C132525143 @default.