论文阅读 [TPAMI-2022] U2Fusion: A Unified Unsupervised Image Fusion Network

论文阅读 [TPAMI-2022] U2Fusion: A Unified Unsupervised Image Fusion Network

论文搜索(studyai.com)

搜索论文: U2Fusion: A Unified Unsupervised Image Fusion Network

搜索论文: http://www.studyai.com/search/whole-site/?q=U2Fusion:+A+Unified+Unsupervised+Image+Fusion+Network

关键字(Keywords)

Image fusion; Task analysis; Feature extraction; Measurement; Supervised learning; Data mining; Training; Image fusion; unified model; unsupervised learning; continual learning

机器学习; 机器视觉

监督学习; 无监督学习; 连续(增量)学习; 图像融合; 多模态感知

摘要(Abstract)

This study proposes a novel unified and unsupervised end-to-end image fusion network, termed as U2Fusion, which is capable of solving different fusion problems, including multi-modal, multi-exposure, and multi-focus cases.

本研究提出了一种新的统一且无监督的端到端图像融合网络，称为U2Fusion，它能够解决不同的融合问题，包括多模态、多曝光和多聚焦情况。.

Using feature extraction and information measurement, U2Fusion automatically estimates the importance of corresponding source images and comes up with adaptive information preservation degrees.

U2Fusion通过特征提取和信息测量，自动估计相应源图像的重要性，并提出自适应信息保留度。.

Hence, different fusion tasks are unified in the same framework.

因此，不同的融合任务统一在同一个框架中。.

Based on the adaptive degrees, a network is trained to preserve the adaptive similarity between the fusion result and source images.

基于自适应度，训练网络以保持融合结果与源图像之间的自适应相似性。.

Therefore, the stumbling blocks in applying deep learning for image fusion, e.g., the requirement of ground-truth and specifically designed metrics, are greatly mitigated.

因此，将深度学习应用于图像融合的障碍，例如，对地面真实度和专门设计的度量的要求，得到了极大的缓解。.

By avoiding the loss of previous fusion capabilities when training a single model for different tasks sequentially, we obtain a unified model that is applicable to multiple fusion tasks.

通过避免在为不同任务顺序训练单个模型时丢失先前的融合能力，我们获得了适用于多个融合任务的统一模型。.

Moreover, a new aligned infrared and visible image dataset, RoadScene (available at https://github.com/hanna-xu/RoadScene), is released to provide a new option for benchmark evaluation.

此外，新的红外和可见光图像数据集RoadScene（可在https://github.com/hanna-xu/RoadScene)，为基准评估提供了一个新选项。.

Qualitative and quantitative experimental results on three typical image fusion tasks validate the effectiveness and universality of U2Fusion.

在三个典型图像融合任务上的定性和定量实验结果验证了U2Fusion的有效性和通用性。.

Our code is publicly available at https://github.com/hanna-xu/U2Fusion…

我们的代码在https://github.com/hanna-xu/U2Fusion…

作者(Authors)

[‘Han Xu’, ‘Jiayi Ma’, ‘Junjun Jiang’, ‘Xiaojie Guo’, ‘Haibin Ling’]