2026
Zero-Shot Image Denoising via Hybrid Prior-Guided Pseudo Sample Generation
CVPR 2026
Zero-shot image denoising has gained prominence in recent years, as it inherently relies on the intrinsic priors of images rather than learning from external data. Nevertheless, most existing methods either fail to fully exploit global priors, or do not properly preserve the fine-grained details gov