通过图6，我们可以发现，训练样本数在100到800时，随着训练赝本数的

By 6, we can see that the number of training samples in the 100-800, with the enhancement of training pseudo this number, more and more high recognition accuracy, with the increase in the number of iterations, the recognition accuracy has also been a slight increase, shows the identification the number of samples with accuracy and depth of neural networks related to the number of iterations.

Through Figure 6, we can find that the number of training samples in 100 to 800, with the increase of the number of training, the accuracy of recognition is higher and higher, with the increase in the number of iterations, the accuracy of identification has also been slightly improved, indicating that the recognition accuracy is related to the number of samples and the number of deep neural network iterations.

Through Figure 6, we can find that when the number of training samples is between 100 and 800, with the increase of the number of training forgeries, the recognition accuracy is getting higher and higher. With the increase of the number of iterations, the recognition accuracy is also slightly improved, which shows that the recognition accuracy is related to the number of samples and the number of iterations of deep neural network.<br>