用语义边缘引导提升扩散模型超分辨率的结构准确性
SCEESR: Semantic-Control Edge Enhancement for Diffusion-Based Super-Resolution
- 引入ControlNet动态控制生成过程中的边缘结构
- 在单次推理中实现高保真度与高效性平衡
- 适合追求快速高质量图像重建的研究者
真实世界图像超分辨率（Real-ISR）需应对复杂退化和固有重建模糊性。尽管生成模型提升了感知质量，但计算成本仍是关键挑战。单步扩散模型虽速度快，却常因蒸馏伪影导致结构失真。为此，我们提出一种新框架，通过控制网络机制引入语义边缘指导，增强单步扩散模型的结构可控性。该方法在单次推断中融合边缘信息，实现动态结构控制。同时设计混合损失函数，结合L2、LPIPS与边缘感知的AME损失，兼顾像素精度、感知质量和几何一致性。实验表明，该方法显著提升结构完整性和真实感，同时保持单步生成的效率，实现输出质量与推理速度的更优平衡。测试数据集将发布于 https://drive.google.com/drive/folders/1amddXQ5orIyjbxHgGpzqFHZ6KTolinJF?usp=drive_link，相关代码见 https://github.com/ARBEZ-ZEBRA/SCEESR。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。