Image classification for user feedback using Deep Learning Techniques

Abstract

In this digital era, automatic human facial expression recognition is considered as an important component in computer vision. Also, it is challenging for machine learning algorithms, where humans can significantly show their expressions. Henceforth, in machine learning methods, deep learning is considered as a novel technology that can classify the images of human faces into different facial expression recognition categories using convolutional neural networks (CNN). In this system, the facial expression recognition is implemented by using CNN network based model with LeNet architecture to improve the prediction of expression results. Here, the proposed research work has utilized a facial expression dataset, which is loaded from Kaggle web resources and this dataset contains seven facial expression tags such as happy, anger, neutral, fear, sad, disgust, and surprise. In this system, along with emotion classifications, gender classification is also merged. Because automatic gender recognition has relevant to the addition of its usages in software applications whereas in social media and social networking websites. With this system, gender and facial expression recognition are explored through face detection using Convolution Neural Network (CNN).The whole motivation behind the work is to improve the way human movement is detected for different legal purposes. The usage of computer vision on the field of customer service, user security, user feedback and many other things. The gender and expression recognition can be used to deal with many real world problems.