Biometric recognition is a challenging research topic with many opportunities due to security and privacy concerns. Biometrics is the authentication technology that deals with the verification and identification of an individual based on behavioral and physiological features, are fingerprints, palm prints, facial features, ears, irises, retinas, signatures, gaits, keystrokes, and voice.
Law enforcement agencies, border control, financial services, various consumer smart devices, Eye movements tracking, Commercial applications, and healthcare are the significant applications of biometrics. In recent history, deep learning techniques have attained state-of-the-art outcomes in various areas, such as computer vision, speech recognition, and natural language processing. In such cases, deep learning models are utilized for biometrics recognition systems to attain improved accuracy and performance. The most impressive applications of deep learning-based biometrics recognition are Face recognition, Ear Recognition, ECG Recognition, Finger-vein Recognition, Fingerprint Recognition, Gait Recognition, Palm Recognition, Signature Recognition, and Iris Recognition.
Biometric recognition system using deep learning models also needs further research in the following research directions: Security and Privacy Issues, Real-time Models for Various Applications, Efficient Memory Models, Biometric Fusion, More Challenging Datasets, Interpretable Deep Models, Few Shot Learning, and Self-Supervised Learning. Numerous literature surveys and reviews have been published on deep learning-based biometric recognition that present effective details such as comprehensive overview, deep learning approaches, strategies, popular datasets, open-ended research problems, performance evaluation parameters, possible future directions, challenges, and future trends.