Invariant feature extraction for facial recognition: A survey of the state-of-the-art
Adam Hassan, Serestina Viriri · 2018
Face recognition systems because of concerned significant application in surveillance and security treatments have assumed an important part in recent years. The existence of an exact balance between the computing cost, robustness and the ability for face recognition is an important characteristic. Further, the ability to design systems achieves the greatest goal of satisfaction encounter different conditions (e.g. aging, illumination, and pose variations) is an interesting and challenging problem. As invariant feature extraction is an important step in the process, we present the two approaches (generative and nongenerative approaches) of invariant feature extraction for facial recognition and discuss the limitations of both. Further, we conclude and suggest the direction of scientific research.