Matches in SemOpenAlex for { <https://semopenalex.org/work/W4382769732> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W4382769732 endingPage "4666" @default.
- W4382769732 startingPage "4666" @default.
- W4382769732 abstract "In classification tasks, such as face recognition and emotion recognition, multimodal information is used for accurate classification. Once a multimodal classification model is trained with a set of modalities, it estimates the class label by using the entire modality set. A trained classifier is typically not formulated to perform classification for various subsets of modalities. Thus, the model would be useful and portable if it could be used for any subset of modalities. We refer to this problem as the multimodal portability problem. Moreover, in the multimodal model, classification accuracy is reduced when one or more modalities are missing. We term this problem the missing modality problem. This article proposes a novel deep learning model, termed KModNet, and a novel learning strategy, termed progressive learning, to simultaneously address missing modality and multimodal portability problems. KModNet, formulated with the transformer, contains multiple branches corresponding to different k-combinations of the modality set S. KModNet is trained using a multi-step progressive learning framework, where the k-th step uses a k-modal model to train different branches up to the k-th combination branch. To address the missing modality problem, the training multimodal data is randomly ablated. The proposed learning framework is formulated and validated using two multimodal classification problems: audio-video-thermal person classification and audio-video emotion classification. The two classification problems are validated using the Speaking Faces, RAVDESS, and SAVEE datasets. The results demonstrate that the progressive learning framework enhances the robustness of multimodal classification, even under the conditions of missing modalities, while being portable to different modality subsets." @default.
- W4382769732 created "2023-07-01" @default.
- W4382769732 creator A5007962043 @default.
- W4382769732 creator A5027960360 @default.
- W4382769732 date "2023-05-11" @default.
- W4382769732 modified "2023-09-26" @default.
- W4382769732 title "Progressive Learning of a Multimodal Classifier Accounting for Different Modality Combinations" @default.
- W4382769732 cites W2017034896 @default.
- W4382769732 cites W2523267570 @default.
- W4382769732 cites W2799429369 @default.
- W4382769732 cites W2803193013 @default.
- W4382769732 cites W2809410879 @default.
- W4382769732 cites W2936134684 @default.
- W4382769732 cites W2997819898 @default.
- W4382769732 cites W3008785076 @default.
- W4382769732 cites W3012721484 @default.
- W4382769732 cites W3026006730 @default.
- W4382769732 cites W3097616280 @default.
- W4382769732 cites W3097741049 @default.
- W4382769732 cites W3104377639 @default.
- W4382769732 cites W3114214226 @default.
- W4382769732 cites W3161402421 @default.
- W4382769732 cites W3175825020 @default.
- W4382769732 cites W3184679245 @default.
- W4382769732 cites W4205720976 @default.
- W4382769732 cites W4307076504 @default.
- W4382769732 cites W4312596733 @default.
- W4382769732 cites W4313164293 @default.
- W4382769732 cites W4361853108 @default.
- W4382769732 cites W4362716276 @default.
- W4382769732 cites W4362721809 @default.
- W4382769732 cites W4365514328 @default.
- W4382769732 cites W4365790697 @default.
- W4382769732 doi "https://doi.org/10.3390/s23104666" @default.
- W4382769732 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37430579" @default.
- W4382769732 hasPublicationYear "2023" @default.
- W4382769732 type Work @default.
- W4382769732 citedByCount "0" @default.
- W4382769732 crossrefType "journal-article" @default.
- W4382769732 hasAuthorship W4382769732A5007962043 @default.
- W4382769732 hasAuthorship W4382769732A5027960360 @default.
- W4382769732 hasBestOaLocation W43827697321 @default.
- W4382769732 hasConcept C119857082 @default.
- W4382769732 hasConcept C144024400 @default.
- W4382769732 hasConcept C153180895 @default.
- W4382769732 hasConcept C154945302 @default.
- W4382769732 hasConcept C199360897 @default.
- W4382769732 hasConcept C2779903281 @default.
- W4382769732 hasConcept C2780226545 @default.
- W4382769732 hasConcept C2780660688 @default.
- W4382769732 hasConcept C36289849 @default.
- W4382769732 hasConcept C41008148 @default.
- W4382769732 hasConcept C63000827 @default.
- W4382769732 hasConcept C9357733 @default.
- W4382769732 hasConcept C95623464 @default.
- W4382769732 hasConceptScore W4382769732C119857082 @default.
- W4382769732 hasConceptScore W4382769732C144024400 @default.
- W4382769732 hasConceptScore W4382769732C153180895 @default.
- W4382769732 hasConceptScore W4382769732C154945302 @default.
- W4382769732 hasConceptScore W4382769732C199360897 @default.
- W4382769732 hasConceptScore W4382769732C2779903281 @default.
- W4382769732 hasConceptScore W4382769732C2780226545 @default.
- W4382769732 hasConceptScore W4382769732C2780660688 @default.
- W4382769732 hasConceptScore W4382769732C36289849 @default.
- W4382769732 hasConceptScore W4382769732C41008148 @default.
- W4382769732 hasConceptScore W4382769732C63000827 @default.
- W4382769732 hasConceptScore W4382769732C9357733 @default.
- W4382769732 hasConceptScore W4382769732C95623464 @default.
- W4382769732 hasIssue "10" @default.
- W4382769732 hasLocation W43827697321 @default.
- W4382769732 hasLocation W43827697322 @default.
- W4382769732 hasLocation W43827697323 @default.
- W4382769732 hasOpenAccess W4382769732 @default.
- W4382769732 hasPrimaryLocation W43827697321 @default.
- W4382769732 hasRelatedWork W2563096758 @default.
- W4382769732 hasRelatedWork W2888540803 @default.
- W4382769732 hasRelatedWork W2904518532 @default.
- W4382769732 hasRelatedWork W2962931510 @default.
- W4382769732 hasRelatedWork W2963386237 @default.
- W4382769732 hasRelatedWork W3200817606 @default.
- W4382769732 hasRelatedWork W4304193542 @default.
- W4382769732 hasRelatedWork W4323572044 @default.
- W4382769732 hasRelatedWork W4380551887 @default.
- W4382769732 hasRelatedWork W4386053843 @default.
- W4382769732 hasVolume "23" @default.
- W4382769732 isParatext "false" @default.
- W4382769732 isRetracted "false" @default.
- W4382769732 workType "article" @default.