Recognition of Human using Periocular Biometrics with Deep Dense Network
DOI : 10.46335/IJIES.2024.9.1.4
Authors : Deepali R. Bhamare , Dr. Pravin S. Patil
Abstract –The COVID-19 pandemic has drastically reduced people's life expectancy and instilled fear in people all around the world. Concerns over the long-term effects of wearing a facemask and social alienation are raised by these requirements, which emphasize the necessity for contactless biometrics for verification—of which ocular biometrics is the best alternative. Generally speaking, biometric feature-based person identification systems are preferred for verifying a person's identity in public places such as ATMs, banks, school attendance systems, airport immigration clearance systems, etc. Compared to other networks like face net, Alexnet, deepiristnet-A, and deepiristnet-B, the UBIPr dataset and hybrid optimum dense capsule network combined with the African vulture algorithm provide superior accuracy for human recognition. It has an error rate that is 3.32 times lower than the other
