FRAMER：频率对齐的自蒸馏与自适应调制，利用扩散先验实现真实世界图像超分辨率
FRAMER: Frequency-Aligned Self-Distillation with Adaptive Modulation Leveraging Diffusion Priors for Real-World Image Super-Resolution
提出一种即插即用的训练方案，利用扩散模型内部特征图的频域分解（LF/HF）进行自蒸馏，通过对比损失和自适应调制增强高频细节重建，可推广到其他基于扩散先验的低层视觉任务。
原文依据 We introduce FRAMER, a plug-and-play training scheme that exploits diffusion priors without changing the backbone or inference.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论真实图像超分辨率（Real-ISR）旨在从具有混合未知退化的低分辨率输入中恢复高分辨率图像。尽管扩散模型在感知质量上超越了GAN，但由于低频偏置和深度方向上“先低频后高频”的层次结构，它们在高频细节的重建上表现不足。我们提出FRAMER，一种即插即用的训练方案，利用扩散先验而不改变骨干网络或推理过程。在每个去噪步骤中，最终层特征图指导所有中间层。教师和学生的特征图通过FFT掩码分解为低频和高频带，以使监督与模型内部的频率层次对齐。对于低频，采用内部对比损失（IntraCL）来稳定全局共享结构。对于高频，采用间对比损失（InterCL），利用随机层和批内负样本锐化实例特定细节。两个自适应调制器——基于频率的自适应权重（FAW）和基于频率的对齐调制（FAM）——重新加权每层的低频/高频信号，并根据当前相似度门控蒸馏。在U-Net和DiT骨干网络（例如Stable Diffusion 2、3）上，FRAMER一致地提升了PSNR/SSIM和感知指标（LPIPS、NIQE、MANIQA、MUSIQ）。消融实验验证了最终层教师和随机层负样本的有效性。代码和项目页面将公开。
英文原文摘要 1340 字符
Real-image super-resolution (Real-ISR) seeks to recover HR images from LR inputs with mixed, unknown degradations. While diffusion models surpass GANs in perceptual quality, they under-reconstruct high-frequency (HF) details due to a low-frequency (LF) bias and a depth-wise "low-first, high-later" hierarchy. We introduce FRAMER, a plug-and-play training scheme that exploits diffusion priors without changing the backbone or inference. At each denoising step, the final-layer feature map teaches all intermediate layers. Teacher and student feature maps are decomposed into LF/HF bands via FFT masks to align supervision with the model's internal frequency hierarchy. For LF, an Intra Contrastive Loss (IntraCL) stabilizes globally shared structure. For HF, an Inter Contrastive Loss (InterCL) sharpens instance-specific details using random-layer and in-batch negatives. Two adaptive modulators, Frequency-based Adaptive Weight (FAW) and Frequency-based Alignment Modulation (FAM), reweight per-layer LF/HF signals and gate distillation by current similarity. Across U-Net and DiT backbones (e.g., Stable Diffusion 2, 3), FRAMER consistently improves PSNR/SSIM and perceptual metrics (LPIPS, NIQE, MANIQA, MUSIQ). Ablations validate the final-layer teacher and random-layer negatives. The code and project page will be publicly released.