Guiding InfoGAN with Semi-Supervision
This work addresses the challenge of improving GAN-based image synthesis with limited labeled data, which is incremental by building on InfoGAN to incorporate semi-supervision.
The paper tackles the problem of learning semantically meaningful and controllable data representations in image synthesis by proposing a semi-supervised GAN architecture (ss-InfoGAN) that leverages few labels (as little as 0.22% of the dataset) to achieve higher quality synthetic samples compared to fully unsupervised settings and speeds up training convergence.
In this paper we propose a new semi-supervised GAN architecture (ss-InfoGAN) for image synthesis that leverages information from few labels (as little as 0.22%, max. 10% of the dataset) to learn semantically meaningful and controllable data representations where latent variables correspond to label categories. The architecture builds on Information Maximizing Generative Adversarial Networks (InfoGAN) and is shown to learn both continuous and categorical codes and achieves higher quality of synthetic samples compared to fully unsupervised settings. Furthermore, we show that using small amounts of labeled data speeds-up training convergence. The architecture maintains the ability to disentangle latent variables for which no labels are available. Finally, we contribute an information-theoretic reasoning on how introducing semi-supervision increases mutual information between synthetic and real data.