{"title":"Recognition of facial expressions using Gaussian based edge direction and texture descriptor","authors":"I. Revina, W. Emmanuel","doi":"10.1109/ICICI.2017.8365292","DOIUrl":null,"url":null,"abstract":"The aim of Facial Expression Recognition (FER) is, based on facial information to observe and realize human emotions. It is an exciting and exigent problem to distinguish the human facial expression and emotion. This paper suggests a Gaussian based Edge Detection and Texture Descriptor (GEDTD) for FER. Regarding 8 Gaussian edge descriptors GEDTD is formed. The proposed GEDTD extract both image texture feature and edge direction. Using Local XOR Coding (LXC) scheme the interior and locality pixels of edge response directions are encoded for extraction. Ultimately these features are combined and it forms the feature vector. The expressions of different poses likely disgust, sad, smile and surprise are trained by using Convolution Neural Network (CNN), which differentiates the facial expressions into disgust, sad, smile and surprise. The suggested process increases the recognition accuracy at an important level. The under taken method is an appropriate one for any recognition requirements.","PeriodicalId":369524,"journal":{"name":"2017 International Conference on Inventive Computing and Informatics (ICICI)","volume":"91 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 International Conference on Inventive Computing and Informatics (ICICI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICICI.2017.8365292","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 1
Abstract
The aim of Facial Expression Recognition (FER) is, based on facial information to observe and realize human emotions. It is an exciting and exigent problem to distinguish the human facial expression and emotion. This paper suggests a Gaussian based Edge Detection and Texture Descriptor (GEDTD) for FER. Regarding 8 Gaussian edge descriptors GEDTD is formed. The proposed GEDTD extract both image texture feature and edge direction. Using Local XOR Coding (LXC) scheme the interior and locality pixels of edge response directions are encoded for extraction. Ultimately these features are combined and it forms the feature vector. The expressions of different poses likely disgust, sad, smile and surprise are trained by using Convolution Neural Network (CNN), which differentiates the facial expressions into disgust, sad, smile and surprise. The suggested process increases the recognition accuracy at an important level. The under taken method is an appropriate one for any recognition requirements.