Face Expression Recognition by Hybrid Local Binary Pattern with Haar Cascade Method
Abstract
This paper presents the idea related to identifying human emotions: anger, happiness, neutral, sad, surprise. We use it to process photographs and videos to recognize human beings, faces, or even handwriting. We use vector to define the image pattern and its different features. The architectures we employed for implementation were the Local Binary Pattern Histogram(LBPH) algorithm for face recognition and Haar-Cascades classifier for face detection. We leveraged joint and transfer learning approaches to produce our best outcomes to further boost our results. In this project, it will tell which expression that the respective image has, with the percentage of the emotion. It is a challenging task for a computer vision to recognize as same as humans through AI. Nowadays, however, Face emotion recognition is more effective and is used in real-time by many Apps for security purposes. Emotion is sensed by us dynamically through the integrated cam and static images from the dataset.
Keywords: Face Detection, Feature Extraction, Image processing, Opencv