Unsupervised Adversarial Image Inpainting
This addresses the problem of generating multiple plausible image reconstructions in data-scarce scenarios for computer vision applications, though it is incremental as it builds on existing GAN methods.
The paper tackles unsupervised image inpainting without paired or unpaired training data, using only incomplete observations and inpainting statistics, and achieves performance comparable to supervised variants on datasets like CelebA, Recipe-1M, and LSUN Bedrooms.
We consider inpainting in an unsupervised setting where there is neither access to paired nor unpaired training data. The only available information is provided by the uncomplete observations and the inpainting process statistics. In this context, an observation should give rise to several plausible reconstructions which amounts at learning a distribution over the space of reconstructed images. We model the reconstruction process by using a conditional GAN with constraints on the stochastic component that introduce an explicit dependency between this component and the generated output. This allows us sampling from the latent component in order to generate a distribution of images associated to an observation. We demonstrate the capacity of our model on several image datasets: faces (CelebA), food images (Recipe-1M) and bedrooms (LSUN Bedrooms) with different types of imputation masks. The approach yields comparable performance to model variants trained with additional supervision.