Generalized Expectation Maximization Framework for Blind Image Super Resolution

TL;DR

This paper introduces SREMN, an end-to-end Bayesian framework for blind image super resolution that integrates learning with the generalized expectation-maximization algorithm, improving accuracy and enabling semi-supervised learning.

Contribution

It presents a novel unified framework combining learning and GEM for blind SISR, reducing error propagation and supporting semi-supervised training.

Findings

Outperforms existing methods in super resolution quality

Demonstrates effectiveness in semi-supervised learning scenarios

Achieves superior results with end-to-end training

Abstract

Learning-based methods for blind single image super resolution (SISR) conduct the restoration by a learned mapping between high-resolution (HR) images and their low-resolution (LR) counterparts degraded with arbitrary blur kernels. However, these methods mostly require an independent step to estimate the blur kernel, leading to error accumulation between steps. We propose an end-to-end learning framework for the blind SISR problem, which enables image restoration within a unified Bayesian framework with either full- or semi-supervision. The proposed method, namely SREMN, integrates learning techniques into the generalized expectation-maximization (GEM) algorithm and infers HR images from the maximum likelihood estimation (MLE). Extensive experiments show the superiority of the proposed method with comparison to existing work and novelty in semi-supervised learning.