Support Vector Machines for Human Face Detection: A Review

Abstract

The computer vision drawback of face detection has over the years become a standard high-requirements benchmark for machine learning ways. Within the last decade, extremely efficient face detection systems are developed that extensively use the character of the image domain to attain correct time period performance. However, the effectiveness of such systems wouldn't be potential while not the progress within the underlying machine learning and classification ways. Now the research area of face recognition technology is much advanced because the research in this area has been conducted for more than 30 years. The main reason for the popularity of face recognition is that it can be used in the different fields like identity authentication, access control and so on. Support vector machine learning may be a comparatively recent methodology that gives a decent generalization performance. Like alternative ways, SVM learning has been applied to the task of face detection, wherever the drawbacks of the technique became evident. Analysis that specializes in accuracy found that competitive performance is feasible however training on adequately giant datasets is difficult. Others tackled the speed issue and whereas varied approximation ways created interactive response times potential, those usually came at a worth of reduced accuracy.