REALTIME EMOTION DETECTION USING KERAS

Computer Engineering

In many automated device applications, such as robotics education, artificial intelligence, and Defence, recognition of facial expressions plays a major role. It is difficult to precisely recognize facial expressions. Approaches to solve the problem of FER (Facial Expression Recognition) can be divided into 1) single static images and 2) image sequences. Similar methods, historically, Researchers used the Multi-layer Perceptron Model, kNearest Neighbours, Help Vector Machines to solve FER. Features such as Local Binary Patterns, Eigenfaces, Facelandmark features, and Texture features were derived from these methods. Among all these strategies, Neural Networks have gained a lot of popularity and are commonly used for Oh. FER. Due to their casual architecture and ability to provide good results without the need for manual feature extraction from raw image data, CNNs (Convolutionary Neural Networks) have recently gained popularity in the field of deep learning. This paper focuses on a study of different CNN-based facial expression recognition techniques. It involves state-of-the-art techniques proposed by various researchers. The paper also illustrates the steps needed for FER to use CNN. This paper also provides an overview of methods focused on CNN and problems that need focus when selecting CNN to solve FER

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.