arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1606.07536cs.CV
收藏

耦合生成对抗网络

Coupled Generative Adversarial Networks

Mitsubishi Electric Research Labs (MERL)(三菱电机研究实验室(MERL))

Ming-Yu Liu, Oncel Tuzel

摘要:

我们提出了耦合生成对抗网络(CoGAN)来学习多域图像的联合分布。与现有的方法相比,这些方法需要训练集中不同域中的对应图像的元组,CoGAN可以在没有任何对应图像元组的情况下学习联合分布。它可以只从边缘分布中提取样本来学习联合分布。这是通过强制执行一个权重共享约束,限制网络容量,有利于联合分布的解决方案,边际分布的产品之一。我们将CoGAN应用于几个联合分布学习任务,包括学习颜色和深度图像的联合分布,以及学习具有不同属性的人脸图像的联合分布。对于每个任务,它成功地学习了联合分布,而没有任何对应图像的元组。我们还展示了它的应用程序的领域自适应和图像变换。

英文摘要:

We propose coupled generative adversarial network (CoGAN) for learning a joint distribution of multi-domain images. In contrast to the existing approaches, which require tuples of corresponding images in different domains in the training set, CoGAN can learn a joint distribution without any tuple of corresponding images. It can learn a joint distribution with just samples drawn from the marginal distributions. This is achieved by enforcing a weight-sharing constraint that limits the network capacity and favors a joint distribution solution over a product of marginal distributions one. We apply CoGAN to several joint distribution learning tasks, including learning a joint distribution of color and depth images, and learning a joint distribution of face images with different attributes. For each task it successfully learns the joint distribution without any tuple of corresponding images. We also demonstrate its applications to domain adaptation and image transformation.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑