Scholarlink

<p>Recently, unsupervised super-resolution has attracted increasing attention due to its promising potential in real-world applications. Early studies proposed generating synthetic low-resolution (LR) images that share similar degradation characteristics with real LR images while preserving the content of high-resolution (HR) images, and utilizing the generated pseudo pairs for supervised training of super-resolution networks. Nevertheless, the SR performance of these approaches is limited by the inevitable domain gap between synthetic and real LR images. In this paper, we propose a novel domain distance aware super-resolution (DASR+) approach with newly designed <b>domain gap aware training</b> strategy, <b>domain distance weighted supervision</b> strategy, and a <b>domain distance adaptive network</b> for unsupervised image super-resolution. Particularly, domain gap aware training takes additional benefit from real LR images and performs feature-level as well as image-level adversarial learning to leverage real-world LR images in the target domain. In addition, domain distance weighted supervision enables a more rational use of generated LR-HR pairs using domain distance information, and the domain distance adaptive SR network injects domain distance information into the network architecture to adaptively adjust the SR mapping. As a result, the SR model trained on synthetic data can be generalized to super-resolve real-world LR images without corresponding ground-truth images. We evaluate DASR+ on both synthetic and real datasets, our approach consistently outperforms state-of-the-art unsupervised SR methods in generating results with more realistic textures.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

DASR+: Training Domain Distance Aware Network for Unsupervised Image Super-Resolution

  • 期刊论文

  • 发表日期：

    2026年5月7日

查看全文
  • Xiaorui Zhao,
  • Yunxuan Wei,
  • Xin Deng,
  • Yawei Li,
  • Radu Timofte,
  • Hengjie Song,
  • Shuhang Gu

摘要

Recently, unsupervised super-resolution has attracted increasing attention due to its promising potential in real-world applications. Early studies proposed generating synthetic low-resolution (LR) images that share similar degradation characteristics with real LR images while preserving the content of high-resolution (HR) images, and utilizing the generated pseudo pairs for supervised training of super-resolution networks. Nevertheless, the SR performance of these approaches is limited by the inevitable domain gap between synthetic and real LR images. In this paper, we propose a novel domain distance aware super-resolution (DASR+) approach with newly designed domain gap aware training strategy, domain distance weighted supervision strategy, and a domain distance adaptive network for unsupervised image super-resolution. Particularly, domain gap aware training takes additional benefit from real LR images and performs feature-level as well as image-level adversarial learning to leverage real-world LR images in the target domain. In addition, domain distance weighted supervision enables a more rational use of generated LR-HR pairs using domain distance information, and the domain distance adaptive SR network injects domain distance information into the network architecture to adaptively adjust the SR mapping. As a result, the SR model trained on synthetic data can be generalized to super-resolve real-world LR images without corresponding ground-truth images. We evaluate DASR+ on both synthetic and real datasets, our approach consistently outperforms state-of-the-art unsupervised SR methods in generating results with more realistic textures.