Deep Reparametrization of Multi-Frame Super-Resolution and Denoising

TL;DR

This paper introduces a deep reparametrization method for multi-frame image restoration that combines classical MAP principles with learned deep features, achieving state-of-the-art results in burst denoising and super-resolution.

Contribution

It presents a novel deep reparametrization of the MAP formulation using learned error metrics and latent image representations, enhancing multi-frame image restoration.

Findings

Sets new state-of-the-art in burst denoising

Achieves superior results in burst super-resolution

Demonstrates the effectiveness of deep reparametrization in image restoration

Abstract

We propose a deep reparametrization of the maximum a posteriori formulation commonly employed in multi-frame image restoration tasks. Our approach is derived by introducing a learned error metric and a latent representation of the target image, which transforms the MAP objective to a deep feature space. The deep reparametrization allows us to directly model the image formation process in the latent space, and to integrate learned image priors into the prediction. Our approach thereby leverages the advantages of deep learning, while also benefiting from the principled multi-frame fusion provided by the classical MAP formulation. We validate our approach through comprehensive experiments on burst denoising and burst super-resolution datasets. Our approach sets a new state-of-the-art for both tasks, demonstrating the generality and effectiveness of the proposed formulation.