Development and Analysis of Biometric Ingress Surveillance

Abstract

 Face recognition is a classic problem in the  field of computer vision and popular due to its wide  applications in real-world problems such as access control,  identity verification, physical security, surveillance, etc.  Recent progress in deep learning techniques and the access  to large-scale face databases has led to a significant  improvement of face recognition accuracy under constrained  and semi-constrained scenarios. Deep neural networks are  shown to surpass human performance on Labelled Face in  the Wild (LFW), which consists of celebrity photos captured  in the wild. This technology can be used to create a  surveillance for the ingress systems with broad scope of  application in multiple scenarios. The major ingress  problems that will be address will include, but are not limited  to: lack of automata, robust face normalization,  discrimination, representation learning and the ambiguity of  facial features caused by information loss. This paper  discusses biometric ingress surveillance. With a brief review  of the subject the paper presents an application in this domain  that highlights the true potential of a surveillance system that  can have face recognition as a biometric authentication to  ingress at homes, offices, universities and other potential  ingress systems.