InfoGAN is a type of generative adversarial network that modifies the GAN objective to encourage it to learn interpretable and meaningful representations. This is done by maximizing the mutual information between a fixed small subset of the GAN’s noise variables and the observations.

Formally, InfoGAN is defined as a minimax game with a variational regularization of mutual information and the hyperparameter $\lambda$:

$$ \min_{G, Q}\max_{D}V_{INFOGAN}\left(D, G, Q\right) = V\left(D, G\right) - \lambda{L}_{I}\left(G, Q\right) $$

Where $Q$ is an auxiliary distribution that approximates the posterior $P\left(c\mid{x}\right)$ - the probability of the latent code $c$ given the data $x$ - and $L_{I}$ is the variational lower bound of the mutual information between the latent code and the observations.

In the practical implementation, there is another fully-connected layer to output parameters for the conditional distribution $Q$ (negligible computation ontop of regular GAN structures). Q is represented with a softmax non-linearity for a categorical latent code. For a continuous latent code, the authors assume a factored Gaussian.

Source: InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

Latest Papers

PAPER DATE
DPD-InfoGAN: Differentially Private Distributed InfoGAN
Vaikkunth MugunthanVignesh GokulLalana KagalShlomo Dubnov
2020-10-22
CiwGAN and fiwGAN: Encoding information in acoustic data to model lexical learning with Generative Adversarial Networks
Gašper Beguš
2020-06-04
Disentanglement based Active Learning
Silpa V SAdarsh KSumitra SRaju K George
2019-12-15
Towards Better Understanding of Disentangled Representations via Mutual Information
Xiaojiang YangWendong BiYitong SunYu ChengJunchi Yan
2019-11-25
Generative Adversarial Networks for Failure Prediction
Shuai ZhengAhmed FarahatChetan Gupta
2019-10-04
Elastic-InfoGAN: Unsupervised Disentangled Representation Learning in Imbalanced Data
Utkarsh OjhaKrishna Kumar SinghCho-Jui HsiehYong Jae Lee
2019-10-01
Generating Geological Facies Models with Fidelity to Diversity and Statistics of Training Images using Improved Generative Adversarial Networks
Lingchen ZhuTuanfeng Zhang
2019-09-23
Maximizing Mutual Information for Tacotron
Peng LiuXixin WuShiyin KangGuangzhi LiDan SuDong Yu
2019-08-30
Unsupervised Classification of Street Architectures Based on InfoGAN
Ning WangXianhan ZengRenjie XieZefei GaoYi ZhengZiran LiaoJunyan YangQiao Wang
2019-05-30
Learning Robotic Manipulation through Visual Planning and Acting
Angelina WangThanard KurutachKara LiuPieter AbbeelAviv Tamar
2019-05-11
IB-GAN: Disentangled Representation Learning with Information Bottleneck GAN
Insu JeonWonkwang LeeGunhee Kim
2019-05-01
Heartbeat Anomaly Detection using Adversarial Oversampling
Jefferson L. P. LimaDavid MacêdoCleber Zanchettin
2019-01-28
Classification of sparsely labeled spatio-temporal data through semi-supervised adversarial learning
Atanas MirchevSeyed-Ahmad Ahmadi
2018-01-26
The Information-Autoencoding Family: A Lagrangian Perspective on Latent Variable Generative Modeling
Shengjia ZhaoJiaming SongStefano Ermon
2018-01-01
Towards Grounding Conceptual Spaces in Neural Representations
Lucas BechbergerKai-Uwe Kühnberger
2017-06-15
beta-VAE: Learning Basic Visual Concepts with a Constrained Variational Framework
| Irina HigginsLoic MattheyArka PalChristopher BurgessXavier GlorotMatthew BotvinickShakir MohamedAlexander Lerchner
2017-04-26
Image Generation and Editing with Variational Info Generative AdversarialNetworks
Mahesh GorijalaAmbedkar Dukkipati
2017-01-17
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
| Xi ChenYan DuanRein HouthooftJohn SchulmanIlya SutskeverPieter Abbeel
2016-06-12

Tasks

TASK PAPERS SHARE
Image Generation 2 22.22%
Active Learning 1 11.11%
Decision Making 1 11.11%
Speech Synthesis 1 11.11%
Visual Tracking 1 11.11%
Anomaly Detection 1 11.11%
Unsupervised Image Classification 1 11.11%
Unsupervised MNIST 1 11.11%

Categories