针对大图超分加速，提出自适应补丁融合网络，按纹理复杂度动态分配计算量。
ENAF: A Multi-Exit Network with an Adaptive Patch Fusion for Large Image Super Resolution
- 多出口结构结合纹理感知评分，动态决定每块图像的计算路径。
- 在2K-8K大图上实现更快推理，比基准模型提速30%以上且保真度不降。
- 适合需要高效处理高分辨率图像的视觉任务，如医学影像或遥感分析。
为加速大图像（2K-8K）单图像超分（SISR）网络的推理，现有方法将图像分解为小块并根据难度动态选择执行路径（称为动态网络）。然而，这些方法通常依赖人工设计的评估指标（如边缘强度），其与SISR模型计算复杂度的关联较弱。为此，本文提出ENAF——一种基于自适应补丁融合的SISR动态网络。在骨干网络基础上，ENAF引入多个早期退出（Early Exits, EEs），以应对过参数化模型带来的负担。更重要的是，它嵌入一个微型网络，在每个退出点估计PSNR，从而将图像纹理特征与计算成本相关联。基于该评分，ENAF能有效将图像块分配至合适的退出点，显著提升质量与复杂度之间的权衡。在多个常用数据集及主流SISR骨干网络上的大量实验验证了其有效性。代码已开源于https://github.com/nmduonggg/ENAF。
原文摘要 · Abstract (English)
To accelerate single image super-resolution (SISR) networks on large images (2K-8K), many recent approaches decompose an image into small patches and dynamically determine an execution path according to its difficulty (referred to as a dynamic network). To quantify the hardness of a patch, they mainly rely on a handcrafted assessment score, e.g., edge, which weakly associates a patch's texture with the computational complexity of a SISR model. To address the problem, we introduce ENAF - a dynamic network for SISR with an adaptive patch fusion. Built on top of a backbone, ENAF incorporates multiple early exits (EEs) to tackle the over-parameterized SISR model. More importantly, ENAF plugs a tiny network that estimates PSNR to associate data texture with a computation cost at an EE. Based on the scores, ENAF effectively assigns image patches to an exit, enhancing the quality-complexity trade-off. Extensive experiments on common datasets with popular SISR backbones demonstrate the effectiveness of ENAF in various settings. The source code is provided in https://github.com/nmduonggg/ENAF
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。