Leverage and discriminant analysis of face recognition
R. Angeline, Krishna Chaitanya Poladi, Rudra Dev Mishra · AIP conference proceedings · 2022
Now a days, security related issue plays a key role ineach and every organization. Every organization consists of internal security methods and makes it work as efficient as possible. One such security method is the online Face recognition process. In recent past days, online examinations are being held in many organizations. Due to this, there is a chance of misusing the system in many ways. One such problem we have identified is that the problem arises between the twoidentical persons / twins. To solve this problem we introduced a new technique where this system cannot be misused with thehelp of transfer learning approach using a CNN model (ResNet) for learning invariant features from the facial images. Human face recognition (HFR) approach provides more challenging issues because of the large intra class variation among the facial images and also due to the less availability of data that needs to be trained to the system for accurate detection. This paper proposes the transfer learning approach for learning invariant features from the facial images more accurately.