51. Face recognition for human identification through integration of complex domain unsupervised and supervised frameworks.
- Author
-
Srivastava, Swati and Sharma, Himanshu
- Abstract
Human identification can be performed through various available biometric traits such as the face, iris, fingerprint, ECG, gait, and ear. Among them, face is one of the most popular and widely used biometrics. In the security domain, early warnings and the trace of suspects can be accomplished using face recognition. The contemplated augmentation projected an intelligent computational model for human recognition which is an ingenious melding of unsupervised outline and complex domain neurocomputing. The unsupervised framework of our proposal constitutes evolutionary fuzzy computations in complex domain. The supervised schema capitalizes on a complex domain neural network with higher-order neurons and resilient propagation algorithm. Trainable multiple stages are populated in this proposal for the estimation of recognition and classification. This proposal offers an intelligent performance on recognition and classification tasks. Comprehensive experimental analysis on the datasets of AR face, PubFig83, and Indian face evidenced the enhanced precision of the proposed model. Our model achieves an impressive accuracy range of 97% to 99% across all datasets. These results clearly demonstrate the superior performance of our approach, showcasing the dominance of the combined unsupervised and supervised frameworks over other state-of-the-art methods. [ABSTRACT FROM AUTHOR]
- Published
- 2024
- Full Text
- View/download PDF