Matches in SemOpenAlex for { <https://semopenalex.org/work/W2971004099> ?p ?o ?g. }
Showing items 1 to 91 of
91
with 100 items per page.
- W2971004099 abstract "We propose a novel method, Modality-based Redundancy Reduction Fusion (MRRF), for understanding and modulating the relative contribution of each modality in multimodal inference tasks. This is achieved by obtaining an (M+1)-way tensor to consider the high-order relationships between M modalities and the output layer of a neural network model. Applying a modality-based tensor factorization method, which adopts different factors for different modalities, results in removing information present in a modality that can be compensated by other modalities, with respect to model outputs. This helps to understand the relative utility of information in each modality. In addition it leads to a less complicated model with less parameters and therefore could be applied as a regularizer avoiding overfitting. We have applied this method to three different multimodal datasets in sentiment analysis, personality trait recognition, and emotion recognition. We are able to recognize relationships and relative importance of different modalities in these tasks and achieves a 1% to 4% improvement on several evaluation measures compared to the state-of-the-art for all three tasks." @default.
- W2971004099 created "2019-09-05" @default.
- W2971004099 creator A5059791595 @default.
- W2971004099 creator A5065856469 @default.
- W2971004099 date "2019-01-01" @default.
- W2971004099 modified "2023-10-17" @default.
- W2971004099 title "Modality-based Factorization for Multimodal Fusion" @default.
- W2971004099 cites W1651753422 @default.
- W2971004099 cites W1963826206 @default.
- W2971004099 cites W1985867508 @default.
- W2971004099 cites W2000215628 @default.
- W2971004099 cites W2029996593 @default.
- W2971004099 cites W2064675550 @default.
- W2971004099 cites W2075398069 @default.
- W2971004099 cites W2077395415 @default.
- W2971004099 cites W2095176743 @default.
- W2971004099 cites W2096391593 @default.
- W2971004099 cites W2099355420 @default.
- W2971004099 cites W2121739212 @default.
- W2971004099 cites W2146334809 @default.
- W2971004099 cites W2154739180 @default.
- W2971004099 cites W2164587673 @default.
- W2971004099 cites W2164699598 @default.
- W2971004099 cites W2184188583 @default.
- W2971004099 cites W22229905 @default.
- W2971004099 cites W2258054274 @default.
- W2971004099 cites W2465534249 @default.
- W2971004099 cites W2470413457 @default.
- W2971004099 cites W2619383789 @default.
- W2971004099 cites W2740550900 @default.
- W2971004099 cites W2772633765 @default.
- W2971004099 cites W2899771611 @default.
- W2971004099 cites W2963128932 @default.
- W2971004099 cites W2963710346 @default.
- W2971004099 cites W2964010806 @default.
- W2971004099 cites W2964121744 @default.
- W2971004099 cites W2964346351 @default.
- W2971004099 doi "https://doi.org/10.18653/v1/w19-4331" @default.
- W2971004099 hasPublicationYear "2019" @default.
- W2971004099 type Work @default.
- W2971004099 sameAs 2971004099 @default.
- W2971004099 citedByCount "13" @default.
- W2971004099 countsByYear W29710040992019 @default.
- W2971004099 countsByYear W29710040992020 @default.
- W2971004099 countsByYear W29710040992021 @default.
- W2971004099 countsByYear W29710040992022 @default.
- W2971004099 countsByYear W29710040992023 @default.
- W2971004099 crossrefType "proceedings-article" @default.
- W2971004099 hasAuthorship W2971004099A5059791595 @default.
- W2971004099 hasAuthorship W2971004099A5065856469 @default.
- W2971004099 hasBestOaLocation W29710040991 @default.
- W2971004099 hasConcept C119857082 @default.
- W2971004099 hasConcept C144024400 @default.
- W2971004099 hasConcept C153180895 @default.
- W2971004099 hasConcept C154945302 @default.
- W2971004099 hasConcept C22019652 @default.
- W2971004099 hasConcept C2776214188 @default.
- W2971004099 hasConcept C2779903281 @default.
- W2971004099 hasConcept C2780226545 @default.
- W2971004099 hasConcept C36289849 @default.
- W2971004099 hasConcept C41008148 @default.
- W2971004099 hasConcept C50644808 @default.
- W2971004099 hasConceptScore W2971004099C119857082 @default.
- W2971004099 hasConceptScore W2971004099C144024400 @default.
- W2971004099 hasConceptScore W2971004099C153180895 @default.
- W2971004099 hasConceptScore W2971004099C154945302 @default.
- W2971004099 hasConceptScore W2971004099C22019652 @default.
- W2971004099 hasConceptScore W2971004099C2776214188 @default.
- W2971004099 hasConceptScore W2971004099C2779903281 @default.
- W2971004099 hasConceptScore W2971004099C2780226545 @default.
- W2971004099 hasConceptScore W2971004099C36289849 @default.
- W2971004099 hasConceptScore W2971004099C41008148 @default.
- W2971004099 hasConceptScore W2971004099C50644808 @default.
- W2971004099 hasLocation W29710040991 @default.
- W2971004099 hasLocation W29710040992 @default.
- W2971004099 hasOpenAccess W2971004099 @default.
- W2971004099 hasPrimaryLocation W29710040991 @default.
- W2971004099 hasRelatedWork W2004831463 @default.
- W2971004099 hasRelatedWork W2110287964 @default.
- W2971004099 hasRelatedWork W2168054807 @default.
- W2971004099 hasRelatedWork W2383394264 @default.
- W2971004099 hasRelatedWork W3125968744 @default.
- W2971004099 hasRelatedWork W4224266612 @default.
- W2971004099 hasRelatedWork W4293261942 @default.
- W2971004099 hasRelatedWork W4307407935 @default.
- W2971004099 hasRelatedWork W4320153225 @default.
- W2971004099 hasRelatedWork W73545470 @default.
- W2971004099 isParatext "false" @default.
- W2971004099 isRetracted "false" @default.
- W2971004099 magId "2971004099" @default.
- W2971004099 workType "article" @default.