Ways of Conditioning Generative Adversarial Networks
Abstract
The GANs are generative models whose random samples realistically reflect natural images. It also can generate samples with specific attributes by concatenating a condition vector into the input, yet research on this field is not well studied. We propose novel methods of conditioning generative adversarial networks (GANs) that achieve state-of-the-art results on MNIST and CIFAR-10. We mainly introduce two models: an information retrieving model that extracts conditional information from the samples, and a spatial bilinear pooling model that forms bilinear features derived from the spatial cross product of an image and a condition vector. These methods significantly enhance log-likelihood of test data under the conditional distributions compared to the methods of concatenation.
Cite
@article{arxiv.1611.01455,
title = {Ways of Conditioning Generative Adversarial Networks},
author = {Hanock Kwak and Byoung-Tak Zhang},
journal= {arXiv preprint arXiv:1611.01455},
year = {2016}
}