How Good Are Deep Generative Models for Solving Inverse Problems?
It addresses the understudied reliability of generative models for inverse problems, which is important for researchers and practitioners in computer vision and image processing, though it is incremental as it focuses on evaluation rather than new methods.
This study evaluated deep generative models (diffusion, GANs, IMLE) on inverse problems like super-resolution and colourization, finding that the IMLE-based CHIMLE method outperformed others in producing valid solutions and reliable uncertainty estimates.
Deep generative models, such as diffusion models, GANs, and IMLE, have shown impressive capability in tackling inverse problems. However, the validity of model-generated solutions w.r.t. the forward problem and the reliability of associated uncertainty estimates remain understudied. This study evaluates recent diffusion-based, GAN-based, and IMLE-based methods on three inverse problems, i.e., $16\times$ super-resolution, colourization, and image decompression. We assess the validity of these models' outputs as solutions to the inverse problems and conduct a thorough analysis of the reliability of the models' estimates of uncertainty over the solution. Overall, we find that the IMLE-based CHIMLE method outperforms other methods in terms of producing valid solutions and reliable uncertainty estimates.