Evaluation of the Performance of Facial Recognition Model Based on Multiple Algorithms

Facial recognition Algorithms are increasingly being used in many industries and it is weighty to understand the algorithms used in this technology and its impact on science and technology, the Convolutional neural network CNN. This article will provide a summary of facial recognition algorithms technology, discuss the algorithms used in this technology, and research the significance of facial recognition technology in science and technology. Moreover, several methods can be used for facial recognition in the humanoid robotic head. One of the most effective methods is the machine learning technique known as deep learning algorithms. Moreover, Deep learning algorithms are used to perform real-time facial recognition using the robot's camera for image processing. Based on the learned feature extraction for recognition during training, the network will predict the feature extraction that represents the recognized face. In recent years, advances have been made in the field of facial recognition, with many algorithms and techniques being developed to improve accuracy and speed. Additionally, the image dataset was divided into two parts: 130 random default photos and 20 photos of known people. Following the training of the network based on these groups of images, the facial features are extracted from the data. After the data is checked, if the data of the recognized image is the same as one of the existing images, the result will be “This person has been identified successfully”. Also, if the result is negative, it means “the person has not been identified”. Here, the best algorithm among 5 face recognition algorithms (PCA, SVM, LDA, LHP, EF) will be selected based on the accuracy and running time of each algorithm. So, based on the calculation results from the training algorithm, the SVM algorithm is chosen to be used for face recognition in the robot head, according to which it is found that the SVM algorithm is better than the rest in terms of accuracy. accurate, but in terms of execution time, it is found that the PCA algorithm takes less time, while its accuracy is lower than SVM. Because of the sensitivity of this process, the SVM algorithm was selected for its high accuracy, since it has a 98% accuracy while the PCA algorithm has a 98% accuracy. It took 1.02 seconds to complete the run with 90.05% accuracy. Finally, the SVM algorithm was selected to apply in the design of the robot head for face recognition