Discriminant face features extraction, analysis its application in multipose face recognization: a survey

摘要

As one of the excellent learning andudclassification performance, SVM and ISVM has become audresearch topic in the field of machine learning and has beenudapplied in many areas, such as face detection andudrecognition, handwriting automatic identification andudautomatic text categorization. Face recognition is audchallenging computer vision problem. Given a faceuddatabase, goal of face recognition is to compare the inputudimage class with all the classes and then declare a decisionudthat identifies to whom the input image class belongs to or ifudit doesn’t belong to the database at all. In this survey, weudstudy face recognition as a pattern classification problem.Inudthis paper, we study the concept of SVM and sophisticatedudclassification techniques for face recognition using the SVMudand ISVM along with the advantages and disadvantages.udThis paper not only provides an up-to-date critical survey ofudmachine learning techniques but also performance analysisudof various SVM and ISVM techniques for face recognitionudare compared.