Generative Adversarial Networks
Generative Adversarial NetworksIt is an unsupervised learning method that is implemented by letting two neural networks compete with each other. This method was proposed by Ian Goodfellow in 2014.
The generative adversarial network consists of a generative network and a discriminative network. The generative network takes random sampling in the latent space as input, and the output results need to imitate the real samples in the training set as much as possible; the input of the discriminative network is the real samples and the output of the generative network. The purpose is to distinguish the output of the generative network from the real samples as much as possible.
The generative network needs to deceive the discriminative network as much as possible. The two networks resist each other and constantly adjust parameters. The ultimate goal is to make the discriminative network unable to judge the authenticity of the results. Generative adversarial networks are often used to generate realistic pictures, and can also be used to generate videos, three-dimensional object models, etc.