Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897725862> ?p ?o ?g. }
Showing items 1 to 80 of
80
with 100 items per page.
- W2897725862 abstract "Emotions are often perceived by humans through a series of multimodal cues, such as verbal expressions, facial expressions and gestures. In order to recognise emotions automatically, reliable emotional labels are required to learn a mapping from human expressions to corresponding emotions. Dimensional emotion models have become popular and have been widely applied for annotating emotions continuously in the time domain. However, the statistical relationship between emotional dimensions is rarely studied. This paper provides a solution to automatic emotion recognition for the Audio/Visual Emotion Challenge (AVEC) 2018. The objective is to find a robust way to detect emotions using more reliable emotion annotations in the valence and arousal dimensions. The two main contributions of this paper are: 1) the proposal of a new approach capable of generating more dependable emotional ratings for both arousal and valence from multiple annotators by extracting consistent annotation features; 2) the exploration of the valence and arousal distribution using outlier detection methods, which shows a specific oblique elliptic shape. With the learned distribution, we are able to detect the prediction outliers based on their local density deviations and correct them towards the learned distribution. The proposed method performance is evaluated on the RECOLA database containing audio, video and physiological recordings. Our results show that a moving average filter is sufficient to remove the incidental errors in annotations. The unsupervised dimensionality reduction approaches could be used to determine a gold standard annotations from multiple annotations. Compared with the baseline model of AVEC 2018, our approach improved the arousal and valence prediction of concordance correlation coefficient significantly to respectively 0.821 and 0.589." @default.
- W2897725862 created "2018-10-26" @default.
- W2897725862 creator A5013038112 @default.
- W2897725862 creator A5022803097 @default.
- W2897725862 creator A5054399837 @default.
- W2897725862 creator A5089206667 @default.
- W2897725862 date "2018-10-15" @default.
- W2897725862 modified "2023-10-18" @default.
- W2897725862 title "Towards a Better Gold Standard" @default.
- W2897725862 cites W1785074626 @default.
- W2897725862 cites W1923565741 @default.
- W2897725862 cites W1975900269 @default.
- W2897725862 cites W1981377085 @default.
- W2897725862 cites W1987048275 @default.
- W2897725862 cites W2005418748 @default.
- W2897725862 cites W2016381774 @default.
- W2897725862 cites W2025905516 @default.
- W2897725862 cites W2039055144 @default.
- W2897725862 cites W2045528981 @default.
- W2897725862 cites W2051387832 @default.
- W2897725862 cites W2052097980 @default.
- W2897725862 cites W2072803611 @default.
- W2897725862 cites W2080898270 @default.
- W2897725862 cites W2085662862 @default.
- W2897725862 cites W2132555391 @default.
- W2897725862 cites W2136153767 @default.
- W2897725862 cites W2144182447 @default.
- W2897725862 cites W2145332605 @default.
- W2897725862 cites W2149628368 @default.
- W2897725862 cites W2153822685 @default.
- W2897725862 cites W2158557735 @default.
- W2897725862 cites W2239141610 @default.
- W2897725862 cites W2296719434 @default.
- W2897725862 cites W2406223855 @default.
- W2897725862 cites W2506591631 @default.
- W2897725862 cites W2530867395 @default.
- W2897725862 cites W2531338168 @default.
- W2897725862 cites W2531466839 @default.
- W2897725862 cites W2531648894 @default.
- W2897725862 cites W2610961739 @default.
- W2897725862 cites W2746419079 @default.
- W2897725862 cites W2765291577 @default.
- W2897725862 cites W2767087747 @default.
- W2897725862 cites W2801195133 @default.
- W2897725862 cites W4252792238 @default.
- W2897725862 doi "https://doi.org/10.1145/3266302.3266307" @default.
- W2897725862 hasPublicationYear "2018" @default.
- W2897725862 type Work @default.
- W2897725862 sameAs 2897725862 @default.
- W2897725862 citedByCount "9" @default.
- W2897725862 countsByYear W28977258622019 @default.
- W2897725862 countsByYear W28977258622020 @default.
- W2897725862 countsByYear W28977258622021 @default.
- W2897725862 crossrefType "proceedings-article" @default.
- W2897725862 hasAuthorship W2897725862A5013038112 @default.
- W2897725862 hasAuthorship W2897725862A5022803097 @default.
- W2897725862 hasAuthorship W2897725862A5054399837 @default.
- W2897725862 hasAuthorship W2897725862A5089206667 @default.
- W2897725862 hasBestOaLocation W28977258622 @default.
- W2897725862 hasConcept C41008148 @default.
- W2897725862 hasConceptScore W2897725862C41008148 @default.
- W2897725862 hasFunder F4320320924 @default.
- W2897725862 hasLocation W28977258621 @default.
- W2897725862 hasLocation W28977258622 @default.
- W2897725862 hasOpenAccess W2897725862 @default.
- W2897725862 hasPrimaryLocation W28977258621 @default.
- W2897725862 hasRelatedWork W2093578348 @default.
- W2897725862 hasRelatedWork W2130043461 @default.
- W2897725862 hasRelatedWork W2350741829 @default.
- W2897725862 hasRelatedWork W2358668433 @default.
- W2897725862 hasRelatedWork W2376932109 @default.
- W2897725862 hasRelatedWork W2382290278 @default.
- W2897725862 hasRelatedWork W2390279801 @default.
- W2897725862 hasRelatedWork W2748952813 @default.
- W2897725862 hasRelatedWork W2899084033 @default.
- W2897725862 hasRelatedWork W3004735627 @default.
- W2897725862 isParatext "false" @default.
- W2897725862 isRetracted "false" @default.
- W2897725862 magId "2897725862" @default.
- W2897725862 workType "article" @default.