An Enhanced Generative Adversarial Network and Target Sample Recognition Method

Abstract

The invention relates to the field of computer application technology, in particular to an enhanced generative confrontation network and a target sample identification method. The enhanced generative confrontation network of the present invention includes at least one enhanced generator and at least one enhanced discriminator, the enhanced generator processes the obtained initial data to obtain generated data, and provides the generated data to the enhanced discriminator, The enhanced discriminator processes the generated data and feeds back the classification result to the enhanced generator. The enhanced discriminator includes a volume base layer, a basic capsule layer, a convolution capsule layer and a classification capsule layer. The volume base layer, The base capsule layer, convolutional capsule layer and classification capsule layer are sequentially connected.

Description

technical field [0001] The invention relates to the technical field of computer applications, in particular to an enhanced generative confrontation network and a target sample identification method. Background technique [0002] Generative Adversarial Networks (GAN, Generative Adversarial Networks) have received extensive attention and applications in unsupervised learning of complex distributions in recent years. A Generative Adversarial Network (GAN) is a deep learning model. The model consists of two modules: the generator model (G) and the discriminator model (D). GAN produces a fairly good output through the mutual game learning of the generator and the discriminator. The generator and discriminator are usually composed of multi-layer networks containing convolutional and / or fully connected layers. The generator and discriminator must be differentiable, but not necessarily directly invertible. The training goal of GAN is to obtain the parameters that maximize the cla...

AI Technical Summary

Problems solved by technology

[0004] The existing technology mainly has the following disadvantages: ① Deep learning models, especially the widely used deep convolutional neural network models, need to rely on a large number of training samples to show their advantages

However, the existing generative confrontation network (GAN) models usually have the following problems: (a) GAN is difficult to train, especially under the condition of complex distribution of multi-category data; (b) GAN is easy to cause model collapse, which leads to failure of the raw model

Images