SCEESR: Semantic-Control Edge Enhancement for Diffusion-Based Super-Resolution

TL;DR

This paper introduces SCEESR, a super-resolution method that enhances one-step diffusion models with semantic edge guidance via ControlNet, improving structural accuracy and realism efficiently.

Contribution

It proposes a novel SR framework combining ControlNet-guided edge enhancement with a hybrid loss for improved structural and perceptual quality in diffusion-based super-resolution.

Findings

Improved structural integrity and realism in super-resolved images.

Maintains efficiency of one-step diffusion models.

Achieves better quality-speed trade-off in super-resolution.

Abstract

Real-world image super-resolution (Real-ISR) must handle complex degradations and inherent reconstruction ambiguities. While generative models have improved perceptual quality, a key trade-off remains with computational cost. One-step diffusion models offer speed but often produce structural inaccuracies due to distillation artifacts. To address this, we propose a novel SR framework that enhances a one-step diffusion model using a ControlNet mechanism for semantic edge guidance. This integrates edge information to provide dynamic structural control during single-pass inference. We also introduce a hybrid loss combining L2, LPIPS, and an edge-aware AME loss to optimize for pixel accuracy, perceptual quality, and geometric precision. Experiments show our method effectively improves structural integrity and realism while maintaining the efficiency of one-step generation, achieving a superior balance between output quality and inference speed. The results of test datasets will be published at https://drive.google.com/drive/folders/1amddXQ5orIyjbxHgGpzqFHZ6KTolinJF?usp=drive_link and the related code will be published at https://github.com/ARBEZ-ZEBRA/SCEESR.