Transitional Learning: Exploring the Transition States of Degradation for Blind Super-resolution

TL;DR

This paper introduces TLSR, a novel end-to-end method for blind super-resolution that models degradation transitions to improve performance and efficiency without iterative inference.

Contribution

The paper proposes a transitional learning framework that effectively models unknown degradations as interpretable transitions, eliminating the need for iterative optimization in blind SR.

Findings

TLSR outperforms state-of-the-art blind SR methods in accuracy.

TLSR achieves lower computational complexity.

The transitional degradation model improves interpretability and robustness.

Abstract

Being extremely dependent on iterative estimation of the degradation prior or optimization of the model from scratch, the existing blind super-resolution (SR) methods are generally time-consuming and less effective, as the estimation of degradation proceeds from a blind initialization and lacks interpretable degradation priors. To address it, this paper proposes a transitional learning method for blind SR using an end-to-end network without any additional iterations in inference, and explores an effective representation for unknown degradation. To begin with, we analyze and demonstrate the transitionality of degradations as interpretable prior information to indirectly infer the unknown degradation model, including the widely used additive and convolutive degradations. We then propose a novel Transitional Learning method for blind Super-Resolution (TLSR), by adaptively inferring a transitional transformation function to solve the unknown degradations without any iterative operations in inference. Specifically, the end-to-end TLSR network consists of a degree of transitionality (DoT) estimation network, a homogeneous feature extraction network, and a transitional learning module. Quantitative and qualitative evaluations on blind SR tasks demonstrate that the proposed TLSR achieves superior performances and costs fewer complexities against the state-of-the-art blind SR methods. The code is available at github.com/YuanfeiHuang/TLSR.