Description of Invention
The 2.5D face recognition system, incorporates depth features from 2.5D data (depth images), improve the accuracy and robustness of face recognition. By utilizing this additional layer of information, 2.5D systems outperform traditional 2D methods. In this project, we propose a 2.5D face recognition system based on two fine-tuned deep learning models: EfficientNet B1 and EfficientNet B4. These models are paired with optimizers to reduce the computational load while ensuring high recognition accuracy. The main goal of this project is to evaluate and compare the performance of these optimizers to find out which optimizer performs the best.