1. InfoGAN: Interpretable Representation Learning by
Information Maximizing Generative Adversarial Nets
Xi Chen1,2, Yan Duan1,2, Rein Houthooft1,2, John Schulman1,2,
Ilya Sutskever2, Pieter Abbeel1,2
@NIPS読み会?関西
2016/11/12
担当者: 大阪大学 堀井隆斗
1 UC Berkeley, Department of Electrical Engineering and Computer Science
2 OpenAI
27. [Goodfellow+, 2014] Ian J. Goodfellow, Jean Pouget-Abadiey, Mehdi Mirza, Bing Xu, David Warde-Farley,
Sherjil Ozairz, Aaron Courville, and Yoshua Bengio, Generative Adversarial
Nets,NIPS2014
[Randford+, 2015] Alec Radford, Luke Metz, and Soumith Chintala, Unsupervised Representation
Learning with Deep Convolutional Generative Adversarial Networks, ICLR 2016
[Kingma+, 2014] Diederik P. Kingm, Danilo J. Rezendey, Shakir Mohamedy, and Max Welling, Semi-
supervised Learning with Deep Generative Models, NIPS2014
[Barber and Agakov, 2003] David Barber and Felix Agakov, The IM Algorithm : A variational approach to
Information Maximization, NIPS2003
Chainer-DCGAN: http://mattya.github.io/chainer-DCGAN/, ChainerによるDCGANのデモ
参考文献
Editor's Notes
The GAN formulation uses a simple factored continuous input noise vector z, while imposing no
restrictions on the manner in which the generator may use this noise. As a result, it is possible that
the noise will be used by the generator in a highly entangled way, causing the individual dimensions
of z to not correspond to semantic features of the data.