Next-Scale Prediction: A Self-Supervised Approach for Real-World Image Denoising

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shan, Yiwen, Zhao, Haiyu, Hu, Peng, Peng, Xi, Gou, Yuanbiao
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917376597950464
author Shan, Yiwen
Zhao, Haiyu
Hu, Peng
Peng, Xi
Gou, Yuanbiao
author_facet Shan, Yiwen
Zhao, Haiyu
Hu, Peng
Peng, Xi
Gou, Yuanbiao
contents Self-supervised real-world image denoising remains a fundamental challenge, arising from the antagonistic trade-off between decorrelating spatially structured noise and preserving high-frequency details. Existing blind-spot network (BSN) methods rely on pixel-shuffle downsampling (PD) to decorrelate noise, but aggressive downsampling fragments fine structures, while milder downsampling fails to remove correlated noise. To address this, we introduce Next-Scale Prediction (NSP), a novel self-supervised paradigm that decouples noise decorrelation from detail preservation. NSP constructs cross-scale training pairs, where BSN takes low-resolution, fully decorrelated sub-images as input to predict high-resolution targets that retain fine details. As a by-product, NSP naturally supports super-resolution of noisy images without retraining or modification. Extensive experiments demonstrate that NSP achieves state-of-the-art self-supervised denoising performance on real-world benchmarks, significantly alleviating the long-standing conflict between noise decorrelation and detail preservation. The code is available at https://github.com/XLearning-SCU/2026-CVPR-NSP.
format Preprint
id arxiv_https___arxiv_org_abs_2512_21038
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Next-Scale Prediction: A Self-Supervised Approach for Real-World Image Denoising
Shan, Yiwen
Zhao, Haiyu
Hu, Peng
Peng, Xi
Gou, Yuanbiao
Computer Vision and Pattern Recognition
Self-supervised real-world image denoising remains a fundamental challenge, arising from the antagonistic trade-off between decorrelating spatially structured noise and preserving high-frequency details. Existing blind-spot network (BSN) methods rely on pixel-shuffle downsampling (PD) to decorrelate noise, but aggressive downsampling fragments fine structures, while milder downsampling fails to remove correlated noise. To address this, we introduce Next-Scale Prediction (NSP), a novel self-supervised paradigm that decouples noise decorrelation from detail preservation. NSP constructs cross-scale training pairs, where BSN takes low-resolution, fully decorrelated sub-images as input to predict high-resolution targets that retain fine details. As a by-product, NSP naturally supports super-resolution of noisy images without retraining or modification. Extensive experiments demonstrate that NSP achieves state-of-the-art self-supervised denoising performance on real-world benchmarks, significantly alleviating the long-standing conflict between noise decorrelation and detail preservation. The code is available at https://github.com/XLearning-SCU/2026-CVPR-NSP.
title Next-Scale Prediction: A Self-Supervised Approach for Real-World Image Denoising
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2512.21038