Lip print‐based identification using traditional and deep learning

The concept of biometric identification is centred around the theory that every individual unique and has distinct characteristics. Various metrics such as fingerprint, face, iris, or retina are adopted for this purpose. Nonetheless, new alternatives needed to establish identity individuals on occasions where above techniques unavailable. One emerging method human recognition lip-based identification. It can be treated a kind measure. patterns found lip permanent unless subjected alternations trauma. Therefore, prints serve purpose confirming an individual's identity. main objective work design experiments using computer vision methods recognise solely based their prints. This article compares traditional deep learning how they perform common dataset first pipeline with Speeded Up Robust Features either SVM K-NN machine classifier, which achieved accuracy 95.45% 94.31%, respectively. A second performance VGG16 VGG19 architectures. approach obtained 91.53% 93.22%,