Face Detection and Classification using Convolutional Neural Network with Channel Attention Module

Automatic face detection is one of the most challenging in computer vision, FER is a wide range of application in human- computer interaction, behavior, and human expression. However, most of these related researches use general image classification network, which lead inadaptability while applying to face detection. This paper proposed a deep learning using Convolution Neural Network (CNN) with Channel Attention Module. The FER 2013 dataset is utilized in this research for effective classification of face detection. The Data Augmentation is used in this experiment for data pre-processing and feature extraction using high feature generation pyramid (HFGP) and low feature generation pyramid (LFGP). Face detector using Single-Short multi biox detector (SSD) and ResNet 10 based face detection. Then, face detector are given imput to channel attention module classifer which utlized Deep Learning (DL) for effective classification. The obtained result show that the proosed DL using CNN model achieves better accuracy of 99.90% on FER 2013 dataset which ensure accurate classification compared to other existing methods like Engagement Index, Deep Neural Net and Zoning based model.