Inverse problem regularization with hierarchical variational autoencoders
This addresses image restoration for natural images of any size, offering an incremental improvement by integrating existing techniques.
The paper tackles ill-posed inverse problems in image restoration by using a hierarchical variational autoencoder as a prior, combining denoiser-based and generative model approaches, and shows competitive results with state-of-the-art methods.
In this paper, we propose to regularize ill-posed inverse problems using a deep hierarchical variational autoencoder (HVAE) as an image prior. The proposed method synthesizes the advantages of i) denoiser-based Plug \& Play approaches and ii) generative model based approaches to inverse problems. First, we exploit VAE properties to design an efficient algorithm that benefits from convergence guarantees of Plug-and-Play (PnP) methods. Second, our approach is not restricted to specialized datasets and the proposed PnP-HVAE model is able to solve image restoration problems on natural images of any size. Our experiments show that the proposed PnP-HVAE method is competitive with both SOTA denoiser-based PnP approaches, and other SOTA restoration methods based on generative models.