Deep learning for facial emotion recognition using custom CNN architecture

Abstract

Human facial expressions are an indication of true emotions. To recognize facial expressions accurately is useful in the field of Artificial Intelligence, Computing, Medical, e-Education, and many more. The facial expression recognition (FER) system detects emotion through facial expression. But, it is challenging to detect facial emotions accurately. However, recent advancements in technology, research, and availability of facial expression datasets have led to the development of many FER systems which can accurately detect facial emotions. Past research in the field of FER indicates With Convolutional Neural Networks (CNNs), deep learning techniques are the most advanced presently. Custom CNN Architecture is used to implement basic facial emotion recognition in static images in this paper. A K-fold cross-validation method was used to train them using FER13, CK+, and the JAFFE data set. On the seven classes of fundamental emotions, including anger, disgust, fear, happiness, neutrality, sorrow, and surprise, the FER13, CK+, and JAFFE datasets had an accuracy rate of 91.58 percent. Given the difficulty of developing unique CNN architecture, this study’s accurate findings contrast well with those of previous studies.

Facial Emotion Recognition in Verbal Communication Based on Deep Learning

“…We also assessed the robustness of our model using the CK+ dataset, where our model achieved a promising result compared to state-of-the-art methods. The proposed model achieved 1.48%, 9.98%, and 3.29% higher accuracy compared with those of Hasani et al [54], Borgalli et al [46], and Bodapati et al [6], respectively. We then assessed the proposed model using the KDEF dataset.…”

“…The datasets were divided into training, testing, and validation data, where 60% of data were selected for training, 20% for testing, and 20% for model validation. We followed a state-of-the-art method to split the dataset between training, testing, and validation [46]. Before choosing these percentages, we also tested the proposed model over several variants of data splitting, meaning that the proposed model could effectively learn with a lower amount of data.…”

“…We conducted several experiments to evaluate the performance of the proposed FER model with other state-of-the-art methods, as shown in Table 9. To evaluate the model's robustness, we first compared the performance of our model with FER-2013; the proposed model surpassed an accuracy of 23.2%, 17.2%, 0.6%, and 2.42%, respectively, compared to the models of Arriaga et al [51], J. Li et al [52], Subramanian et al [53], and Borgalli et al [46]. We also assessed the robustness of our model using the CK+ dataset, where our model achieved a promising result compared to state-of-the-art methods.…”

“…Arriaga et al [51] Mini-Xception 66.0 J. Li et al [52] CNN with Transfer Learning 72.0 Subramanian et al [53] Three Layer CNN architecture 88.6 Borgalli et al [46] Six We evaluated the performance of a proposed model in real time to compute the processing time of the proposed model over GPU, CPU and resource-constrained device (Jetson Nano). Jetson Nano is a small and powerful computer that runs multiple CNNs in parallel for different applications, such as recognition, segmentation, object detection, and speech processing.…”

DLSANet: Facial expression recognition with double‐code LBP‐layer spatial‐attention network

“…The proposed method achieved good recognition results on both the RAF-DB and the AffectNet datasets. Borgalli and Surve [45] constructed a new CNN model for FER. First, they used the Haar Cascade algorithm to detect faces and removed irrelevant background information.…”