Multimodal Authentication System Based on Audio-Visual Data: A Review

Abstract

Security has become a major concern in this era of fast-growing applications. In such an authentication scheme, the multimodal verification system is crucial. An essential multifunctional biometric technology for enhancing security is access control via speech and visual features. The highly secure multimodal biometric system can be enhanced using voice and visual feature. The drawbacks of single modality evidence is they have limited performance in both security and robustness. This can be resolved by merging knowledge across two modalities, which strengthens the legitimacy and integrity of the evidence. The credibility and reliability of the system will be improved by combining knowledge from two modalities. The system uses speech and face recognition as a biometric approach for figuring out an man or woman, using comparing virtual image or video person and speech and Facial remembrance is a biometric way of detecting a human by matching a live digital picture or video footage. This research examined recent advancements in multimodal identification systems based on a presenter's auditory and visual input. © 2022 IEEE.