Matches in SemOpenAlex for { <https://semopenalex.org/work/W3143658568> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W3143658568 endingPage "105" @default.
- W3143658568 startingPage "96" @default.
- W3143658568 abstract "Emotion is precious, useful for many applications such as public opinion detection and psychological disease prediction. Emotion recognition on multimodal data has attracted extensive attention. Although modifying model structure or multimodal feature fusion methods have contributed a lot to emotion recognition, little attention is paid to mining implicit emotion relationship. In this article, implicit emotion relationship consists of emotion distribution, confusion, and transfer. Emotion distribution allows multiple emotions in one sample, while confusion and transfer imply the prediction confusion and bias. In order to mine implicit emotion relationship in multimodal data, this article employs three image and two text classification models to recognize emotions, respectively. Two prediction emotion synthesis methods (optimal prediction emotion synthesis and majority prediction emotion synthesis) are proposed to synthesize the outputs of multiple models. Based on the results of two emotion synthesis methods, emotion distribution on samples is obtained. Emotion confusion and transfer among different emotion samples are analyzed by relative entropy and Jensen–Shannon divergence. Implicit emotion relationship mining has potential not only in the interpretation of model performance, but also in guiding the development of emotion recognition as prior knowledge. Finally, we take topic scenario as an instance to mine implicit emotion relationships." @default.
- W3143658568 created "2021-04-13" @default.
- W3143658568 creator A5003759585 @default.
- W3143658568 creator A5024801841 @default.
- W3143658568 creator A5036388641 @default.
- W3143658568 creator A5049078993 @default.
- W3143658568 creator A5067265615 @default.
- W3143658568 creator A5080285200 @default.
- W3143658568 date "2021-04-01" @default.
- W3143658568 modified "2023-10-16" @default.
- W3143658568 title "Implicit Emotion Relationship Mining Based on Optimal and Majority Synthesis From Multimodal Data Prediction" @default.
- W3143658568 cites W1523719202 @default.
- W3143658568 cites W1907775068 @default.
- W3143658568 cites W2072787788 @default.
- W3143658568 cites W2147278762 @default.
- W3143658568 cites W2194775991 @default.
- W3143658568 cites W2546875627 @default.
- W3143658568 cites W2618530766 @default.
- W3143658568 cites W2790763247 @default.
- W3143658568 cites W2900431199 @default.
- W3143658568 cites W2985076077 @default.
- W3143658568 cites W2995479270 @default.
- W3143658568 cites W2996799978 @default.
- W3143658568 cites W3001529617 @default.
- W3143658568 cites W4249142469 @default.
- W3143658568 doi "https://doi.org/10.1109/mmul.2021.3071495" @default.
- W3143658568 hasPublicationYear "2021" @default.
- W3143658568 type Work @default.
- W3143658568 sameAs 3143658568 @default.
- W3143658568 citedByCount "7" @default.
- W3143658568 countsByYear W31436585682022 @default.
- W3143658568 countsByYear W31436585682023 @default.
- W3143658568 crossrefType "journal-article" @default.
- W3143658568 hasAuthorship W3143658568A5003759585 @default.
- W3143658568 hasAuthorship W3143658568A5024801841 @default.
- W3143658568 hasAuthorship W3143658568A5036388641 @default.
- W3143658568 hasAuthorship W3143658568A5049078993 @default.
- W3143658568 hasAuthorship W3143658568A5067265615 @default.
- W3143658568 hasAuthorship W3143658568A5080285200 @default.
- W3143658568 hasConcept C119857082 @default.
- W3143658568 hasConcept C124101348 @default.
- W3143658568 hasConcept C154945302 @default.
- W3143658568 hasConcept C204321447 @default.
- W3143658568 hasConcept C23123220 @default.
- W3143658568 hasConcept C2777438025 @default.
- W3143658568 hasConcept C41008148 @default.
- W3143658568 hasConcept C67186912 @default.
- W3143658568 hasConcept C77088390 @default.
- W3143658568 hasConceptScore W3143658568C119857082 @default.
- W3143658568 hasConceptScore W3143658568C124101348 @default.
- W3143658568 hasConceptScore W3143658568C154945302 @default.
- W3143658568 hasConceptScore W3143658568C204321447 @default.
- W3143658568 hasConceptScore W3143658568C23123220 @default.
- W3143658568 hasConceptScore W3143658568C2777438025 @default.
- W3143658568 hasConceptScore W3143658568C41008148 @default.
- W3143658568 hasConceptScore W3143658568C67186912 @default.
- W3143658568 hasConceptScore W3143658568C77088390 @default.
- W3143658568 hasFunder F4320321001 @default.
- W3143658568 hasIssue "2" @default.
- W3143658568 hasLocation W31436585681 @default.
- W3143658568 hasOpenAccess W3143658568 @default.
- W3143658568 hasPrimaryLocation W31436585681 @default.
- W3143658568 hasRelatedWork W2101955803 @default.
- W3143658568 hasRelatedWork W2144190808 @default.
- W3143658568 hasRelatedWork W2151447942 @default.
- W3143658568 hasRelatedWork W2154103141 @default.
- W3143658568 hasRelatedWork W2366644548 @default.
- W3143658568 hasRelatedWork W2376314740 @default.
- W3143658568 hasRelatedWork W2384888906 @default.
- W3143658568 hasRelatedWork W2961085424 @default.
- W3143658568 hasRelatedWork W4306674287 @default.
- W3143658568 hasRelatedWork W4224009465 @default.
- W3143658568 hasVolume "28" @default.
- W3143658568 isParatext "false" @default.
- W3143658568 isRetracted "false" @default.
- W3143658568 magId "3143658568" @default.
- W3143658568 workType "article" @default.