Inverse Problems with Diffusion Models: A MAP Estimation Perspective
This work addresses a key bottleneck in applying diffusion models to inverse problems for computer vision applications, offering a novel optimization perspective that could improve performance in tasks such as inpainting and deblurring.
The paper tackles the challenge of solving inverse problems like image restoration using pre-trained diffusion models without task-specific training, by proposing a MAP estimation framework that models the reverse conditional generation process as an optimization with a tractable gradient, and validates it with extensive experiments across multiple datasets and tasks.
Inverse problems have many applications in science and engineering. In Computer vision, several image restoration tasks such as inpainting, deblurring, and super-resolution can be formally modeled as inverse problems. Recently, methods have been developed for solving inverse problems that only leverage a pre-trained unconditional diffusion model and do not require additional task-specific training. In such methods, however, the inherent intractability of determining the conditional score function during the reverse diffusion process poses a real challenge, leaving the methods to settle with an approximation instead, which affects their performance in practice. Here, we propose a MAP estimation framework to model the reverse conditional generation process of a continuous time diffusion model as an optimization process of the underlying MAP objective, whose gradient term is tractable. In theory, the proposed framework can be applied to solve general inverse problems using gradient-based optimization methods. However, given the highly non-convex nature of the loss objective, finding a perfect gradient-based optimization algorithm can be quite challenging, nevertheless, our framework offers several potential research directions. We use our proposed formulation to develop empirically effective algorithms for image restoration. We validate our proposed algorithms with extensive experiments over multiple datasets across several restoration tasks.