Method for realizing facial expression migration of cyclic generative adversarial network based on spectrum normalization

Abstract

The invention discloses a method for realizing facial expression migration of a cyclic generative adversarial network based on spectrum normalization, which comprises the following steps of: S1, collecting various facial expression pictures, and classifying the facial expression pictures one by one according to facial expressions; S2, preprocessing the pictures, removing blurred pictures, obtaining five key points of the human face by using a human face detection algorithm, and uniformly cutting face pictures according to the key points; S3, constructing a cyclic generative adversarial networkconsisting of a generator and a discriminator, respectively inputting the two types of preprocessed pictures into the network to calculate a loss function, and training the loss function; and S4, obtaining the trained generator as a tool for human face expression migration, and applying the trained generator to actual measurement. The cyclic generative adversarial network based on spectrum normalization can enable one generator to realize migration of multiple facial expressions, and the generated facial expressions can be more natural and have better robustness.

Description

technical field [0001] The invention relates to a realization method of human facial expression migration based on a spectral normalization cycle generation confrontation network, which can be used in the technical field of image processing in computer vision. Background technique [0002] In recent years, with the rapid development of artificial intelligence, deep learning has also become a popular research field, and the proposal of generative confrontation network has accelerated the process of deep learning. Scholars such as Ian Goodfellow of the University of Montreal proposed the generative confrontation network in 2014. In recent years, the generative confrontation network has become one of the research hotspots in deep learning. [0003] A Generative Adversarial Network is a generative model whose structure is inspired by two-person zero-sum games. Generative adversarial networks consist of a generator and a discriminator. The generator can learn from the latent di...

AI Technical Summary

Problems solved by technology

The cyclic generative confrontation network is suitable for dealing with image style transfer, but it is easy to change the overall color of the image when it is applied to the process of local feature transfer of the image such as face removal glasses, that is, changes in other areas of the face may occur after the face removal glasses

Images