An Image Authentication Method Based on Deep Convolutional Generative Adversarial Network

Abstract

In the past, there has been a lot of research in the development of digital watermark technology in image authentication. In this paper, we will propose a Deep Convolutional Generative Adversarial Network (DCGAN) architecture extended by the Generative Adversarial Network (GAN) to conduct training and verification evaluation through three subnets. Try to use a deep learning-based architecture for image authentication. Through the training of the model, the features of the authentication image can be more accurately extracted and the tampered local features can be found too.