发表机构
University of California, San Francisco; School of Mathematical Sciences, Dalian University of Technology; School of Computer Science and Engineering, Nanjing University of Science and Technology; School of Computing and Communications, Lancaster University（加利福尼亚大学旧金山分校; 大连理工大学数学科学学院; 南京理工大学计算机科学与工程学院; 兰卡斯特大学计算与通信学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究提出通用Transformer框架UHDformer++，通过4个协同学习空间及FR-CMT、ACM模块实现超高清图像复原，参数至少减少86%，在5项UHD复原任务上性能显著提升。
AI 中文摘要
我们提出UHDformer++，这是一种基于Transformer的通用框架，用于解决多种超高清（UHD）图像复原任务。UHDformer++在4个协同学习空间中运行：1）高分辨率空间（HR），用于多级特征提取；2）低分辨率空间（LR），用于学习紧凑的代表性特征；3）超分辨率空间（SR），用于对来自SR的低分辨率特征进行上采样；4）高低融合与重建空间（LHFR），用于最终的图像复原。具体而言，HR提取多尺度高分辨率特征并将其与低分辨率线索融合以生成残差图像，而LR则从HR中提炼互补表示以提升复原质量。为向LHFR提供更丰富的特征，SR在融合前对LR输出进行超分辨率处理。我们进一步引入两个模块以衔接高低分辨率空间：特征精细化相关匹配变换（FR-CMT）模块从最大池化与平均池化的高分辨率特征的融合结果中选取前C/r个通道（C表示通道数，r≥1控制压缩程度），以替换低分辨率Transformer中信息含量较低的通道；自适应通道调制器（ACM）自适应地重新校准多尺度高分辨率特征，确保仅任务相关的信息传递至LR。大量实验表明，UHDformer++与近期的最先进方法相比，模型参数至少减少86%，同时在5项UHD复原任务（包括低光图像增强、去雾、去模糊、去雨和去雪）上取得了显著的性能提升。代码将发布在该https URL。
英文摘要
We propose UHDformer++, a general Transformer-based framework to solve numerous Ultra-High-Definition (UHD) image restoration tasks. UHDformer++ operates across $4$ coordinated learning spaces: 1) a high-resolution space (HR) for multi-level feature extraction, 2) a low-resolution space (LR) for learning compact, representative features, 3) a super-resolution space (SR) for upsampling low-resolution features from SR, and 4) a low-high fusion and reconstruction space (LHFR) for final image restoration. Specifically, HR extracts multi-scale high-resolution features and fuses them with low-resolution cues to produce residual images, while LR distills complementary representations from HR to improve restoration quality. To supply LHFR with richer features, SR super-resolves LR outputs before fusion. We further introduce two modules to bridge the high- and low-resolution spaces. The Feature-Refined Correlation Matching Transformation (FR-CMT) module selects the top $C/r~(C~\text{denotes the number of channels;~}r\geq1~\text{controls the squeezing level})$, from the fusion between max- and mean-pooled high-resolution features to replace less informative channels in the low-resolution Transformer. The Adaptive Channel Modulator (ACM) adaptively recalibrates multi-scale high-resolution features, ensuring that only task-relevant information propagates to LR. Extensive experiments demonstrate that UHDformer++ reduces model parameters by at least 86\% compared with recent state-of-the-art methods while achieving substantial performance gains across $5$ UHD restoration tasks, including low-light image enhancement, dehazing, deblurring, deraining, and desnowing. Code will be released at https://github.com/supersupercong/uhdformerplus.
