Facial expression recognition is the paramount segment of non-verbal communication and one frequent procedures of human communication. Facial expression recognition, as one of the foremost evolution regulations of human-computer interaction, enhances the fluency and accuracy of relationship. Over the past few years, deep learning using feature selection of face image data based on machine and deep learning has become increasingly in demand, it has achieved progress, however due to different facial expression and attaining accuracy still remains major issue to be focused. Their accessibility shoots from their potentiality to select righteous features from face image data, via feature selection to achieve high performance at minimum time. In this work a method called, Laplacian Non-linear Logistic Regression and Gravitational Deep Learning (LNLR-GDL) for facial expression recognition is proposed. The proposed method is split into three sections, namely, preprocessing, feature selection and classification. In the first section, preprocessing is conducted with the face recognition dataset where noise reduced preprocessed face images are obtained by employing Unsharp Masking Laplacian Non-linear Filter model. Second with the preprocessed face images, computationally efficient relevant features are selected by means of Logistic Stepwise Regression-based feature selection model. Finally, Gravitational Deep Neural Classification model is applied to the selected features for robust recognition of facial expressions. The proposed LNLR-GDL method along with two state-of-the-art methods is implemented on corrective re-annotation of images for facial expression recognition. The databases comprise of consist of seven emotions viz., anger, contempt, disgust, fear, happiness, neutrality, sadness and surprise. The proposed method is compared with existing methods using three evaluation metrics namely, facial expression recognition accuracy, facial expression recognition time and PSNR. The obtained results demonstrate that the proposed LNLR-GDL method outperforms the state-of-the-art methods.

No abstract

No abstract