Generative Adversarial Networks (GANs) are a well-known technique that is trained on samples (e.g. pictures of fruits) and which after training is able to generate realistic new samples. Conditional GANs (CGANs) additionally provide label information for subclasses (e.g. apple, orange, pear) which enables the GAN to learn more easily and increase the quality of its output samples. We use GANs to learn architectural features of major cities and to generate images of buildings which do not exist. We show that currently available GAN and CGAN architectures are unsuited for this task and propose a custom architecture and demonstrate that our architecture has superior performance for this task and verify its capabilities with extensive experiments.
[1]
Simon Osindero,et al.
Conditional Generative Adversarial Nets
,
2014,
ArXiv.
[2]
Yoshua Bengio,et al.
Generative Adversarial Networks
,
2014,
ArXiv.
[3]
Jitendra Malik,et al.
Generic 3D Representation via Pose Estimation and Matching
,
2016,
ECCV.
[4]
Jon Gauthier.
Conditional generative adversarial nets for convolutional face generation
,
2015
.
[5]
Soumith Chintala,et al.
Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
,
2015,
ICLR.