Dual-discriminator GAN: A GAN way of profile face recognition
Xinyu Zhang, Yang Zhao, Hao Zhang · 2020
A wealth of angle problems occur when facial recognition is performed. At present, the feature extraction network presents eigenvectors with large differences between the frontal face and profile face recognition of the same person in many cases. For this reason, the state-of-the-art facial recognition network have a relatively poor performance on profile faces. In this paper, we proposed a method of image-to-image generating frontal faces from profile faces based on Generative Adversarial Network (GAN). By adopting a new dual discriminator Gan structure and an artificial programed training process, the authority and identity of the generated frontal faces have been greatly improved.