Diffusion Prior-Based Amortized Variational Inference for Noisy Inverse Problems
This addresses scalability and generalization issues in inverse problems for researchers and practitioners in computer vision and image processing, though it is incremental as it builds on existing diffusion prior methods.
The paper tackles the computational inefficiency and lack of generalization in using diffusion models for inverse problems by proposing DAVI, which learns a function to map measurements to posterior distributions, enabling single-step sampling and achieving superior performance in image restoration tasks like Gaussian deblur and super-resolution.
Recent studies on inverse problems have proposed posterior samplers that leverage the pre-trained diffusion models as powerful priors. These attempts have paved the way for using diffusion models in a wide range of inverse problems. However, the existing methods entail computationally demanding iterative sampling procedures and optimize a separate solution for each measurement, which leads to limited scalability and lack of generalization capability across unseen samples. To address these limitations, we propose a novel approach, Diffusion prior-based Amortized Variational Inference (DAVI) that solves inverse problems with a diffusion prior from an amortized variational inference perspective. Specifically, instead of separate measurement-wise optimization, our amortized inference learns a function that directly maps measurements to the implicit posterior distributions of corresponding clean data, enabling a single-step posterior sampling even for unseen measurements. Extensive experiments on image restoration tasks, e.g., Gaussian deblur, 4$\times$ super-resolution, and box inpainting with two benchmark datasets, demonstrate our approach's superior performance over strong baselines. Code is available at https://github.com/mlvlab/DAVI.