| Autor: |
Ibrahim, Dyala R., Shehadeh, Hisham A., Aladaileh, Mohammad A., Alieyan, Kamal, Jaradat, Ghaith M., Telfah, We'am, Wang, Xiaopeng |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings; 2023, Vol. 2979 Issue 1, p1-8, 8p |
| Abstrakt: |
This paper introduces a new "Facial Recognition (FR)" approach based on a new relatively optimization method. This new FR approach is used to mitigate the limitations of recent FR approaches, which are low accuracy and efficiency that caused by high dimensionality of facial images according to unconstrained conditions, such as expression, occlusion, and noise. The "Grasshopper Optimization Algorithm (GOA)" is selected in this work, which outperformed thirteen optimization methods compared early in the literature. In this paper, the efficacy of a hybrid extraction method in conjunction with the K-nearest neighbor classifier is investigated. When compared to other recently developed FR approaches, the performance evaluation of the hybrid "Facial Recognition System (FRS)" depends on the "Grasshopper Optimization Algorithm (GOA)" indicates a near-ideal recognition rate with optimization. In comparisons with the other approaches of recent proposed facial recognition techniques, the proposed algorithm achieved a significant accuracy. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
