{"title":"基于唤醒效价情绪模型和深度学习方法的面部表情识别","authors":"Yong Yang, Yue Sun","doi":"10.1109/icomssc45026.2018.8941829","DOIUrl":null,"url":null,"abstract":"The traditional facial emotion recognition method is classifying basic emotions. But, basic emotions theory is limited to express subtle and disparate emotion. So this paper uses the arousal-valence continuous emotion space model, which can enrich emotion expression. The arousal reflects emotional intensity, and the valence indicates positive and negative emotion. The arousal and valence all have the value in the same range, which is between -1 and 1. In the experiments, it uses convolutional neural network (CNN) in the pre-trained models and support vector regression(SVR). In this model, CNN works as a trained feature extractor and SVR is adopted to train and predict the values of the arousal and valence. Through the predicted values it can be predicted the facial emotion. The contrast experimental results show that the proposed method can get better recognition result than the traditional methods.","PeriodicalId":332213,"journal":{"name":"2018 International Computers, Signals and Systems Conference (ICOMSSC)","volume":"6 3","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":"{\"title\":\"Facial Expression Recognition Based on Arousal- Valence Emotion Model and Deep Learning Method\",\"authors\":\"Yong Yang, Yue Sun\",\"doi\":\"10.1109/icomssc45026.2018.8941829\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The traditional facial emotion recognition method is classifying basic emotions. But, basic emotions theory is limited to express subtle and disparate emotion. So this paper uses the arousal-valence continuous emotion space model, which can enrich emotion expression. The arousal reflects emotional intensity, and the valence indicates positive and negative emotion. The arousal and valence all have the value in the same range, which is between -1 and 1. In the experiments, it uses convolutional neural network (CNN) in the pre-trained models and support vector regression(SVR). In this model, CNN works as a trained feature extractor and SVR is adopted to train and predict the values of the arousal and valence. Through the predicted values it can be predicted the facial emotion. The contrast experimental results show that the proposed method can get better recognition result than the traditional methods.\",\"PeriodicalId\":332213,\"journal\":{\"name\":\"2018 International Computers, Signals and Systems Conference (ICOMSSC)\",\"volume\":\"6 3\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2018 International Computers, Signals and Systems Conference (ICOMSSC)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/icomssc45026.2018.8941829\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 International Computers, Signals and Systems Conference (ICOMSSC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/icomssc45026.2018.8941829","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Facial Expression Recognition Based on Arousal- Valence Emotion Model and Deep Learning Method
The traditional facial emotion recognition method is classifying basic emotions. But, basic emotions theory is limited to express subtle and disparate emotion. So this paper uses the arousal-valence continuous emotion space model, which can enrich emotion expression. The arousal reflects emotional intensity, and the valence indicates positive and negative emotion. The arousal and valence all have the value in the same range, which is between -1 and 1. In the experiments, it uses convolutional neural network (CNN) in the pre-trained models and support vector regression(SVR). In this model, CNN works as a trained feature extractor and SVR is adopted to train and predict the values of the arousal and valence. Through the predicted values it can be predicted the facial emotion. The contrast experimental results show that the proposed method can get better recognition result than the traditional methods.