A Kinship Authentication Algorithm Based on Deep Convolutional Neural Networks
Yichen Hu, Hujun Geng · 2023
The method based on deep learning can extract the nonlinear features of the face through the neural network and then make decisions at the classification layer. This is a typical end-to-end convolutional neural network model, which eliminates the complex image preprocessing stage. Then, pairs of tagged training data were input into the network for iteration to optimize the parameters of the training depth network model. The trained deep network model is used to classify the input test images, and the final accuracy rate is obtained by statistics. The experiment verifies that the improved network based on Siamese network can be effectively applied to the kinship identity authentication algorithm to carry out the task of face recognition comparison, with high accuracy, and effectively narrow the scope of kinship authentication.