A survey of hybrid deep learning-based person identification system using multimodal Palmprint, Ear, Iris, and Face Biometric Features

Abstract

The use of Biometric Systems (BS) as a tool for allowing access to diverse applications has increased recently. Using sensors, an IoT-based multimodal-biometric system combines different biometric models for user verification and authorization. Conventional biometric systems are vulnerable to a range of harmful threats and privacy violations, putting the users who have registered with them in grave danger. These security violations typically target a centralized database where the user's biometric information is kept. The accessibility of highly customized information in these databases significantly raises the privacy problems connected with biometric-aided models. In the modern digital age, biometric technologies are used to provide security and control access because of their one-of-a-kind nature. Biometric systems that include a wide variety of characteristics have been developed as a result of the growing need for biometric technology on a global scale. One of the most important factors in determining robustness is the ability to extract relevant information from a single biometric characteristic. One school of thinking holds that multimodal biometric security systems are superior to unimodal ones in terms of effectiveness and safety. Even with the most advanced multi-biometric architecture, it is still possible for an adversary to get access to the system by exploiting biometric data that has been stolen or hacked.