{"ID":2851532,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2510.19272","arxiv_id":"2510.19272","title":"SCEESR: Semantic-Control Edge Enhancement for Diffusion-Based 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.","short_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 distill...","url_abs":"https://arxiv.org/abs/2510.19272","url_pdf":"https://arxiv.org/pdf/2510.19272v1","authors":"[\"Yun Kai Zhuang\"]","published":"2025-10-22T06:06:01Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"Diffusion Model\"]","project_urls":"[\"https://drive.google.com/drive/folders/1amddXQ5orIyjbxHgGpzqFHZ6KTolinJF?usp=drive_link\"]","has_code":false,"code_links":[{"ID":607911,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2851532,"paper_url":"https://arxiv.org/abs/2510.19272","paper_title":"SCEESR: Semantic-Control Edge Enhancement for Diffusion-Based Super-Resolution","repo_url":"https://github.com/ARBEZ-ZEBRA/SCEESR","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}