arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2003.01947eess.IVcs.CV
收藏

ADRN:基于注意力的深度残差网络高光谱图像去噪

ADRN: Attention-based Deep Residual Network for Hyperspectral Image Denoising

Harbin Institute of Technology(哈尔滨工业大学)

Yongsen Zhao, Deming Zhai, Junjun Jiang, Xianming Liu

摘要:

高光谱图像去噪对于高光谱图像的分类、解译等后续应用具有重要意义。在本文中,我们提出了一种基于注意力的深度残差网络,可以直接学习从噪声HSI到干净HSI的映射。为了联合利用空间-频谱信息,当前频带及其$K$相邻频带被同时利用作为输入。然后,我们采用不同滤波器大小的卷积层来融合多尺度特征,并使用快捷连接来合并多层次信息以更好地去除噪声。此外,采用通道注意机制,使网络集中在最相关的辅助信息和最有利于去噪过程的特征。为了简化训练过程,我们通过残差模式而不是直接预测来重建输出。实验结果表明,我们提出的ADRN计划优于国家的最先进的方法在定量和视觉评价。

英文摘要:

Hyperspectral image (HSI) denoising is of crucial importance for many subsequent applications, such as HSI classification and interpretation. In this paper, we propose an attention-based deep residual network to directly learn a mapping from noisy HSI to the clean one. To jointly utilize the spatial-spectral information, the current band and its $K$ adjacent bands are simultaneously exploited as the input. Then, we adopt convolution layer with different filter sizes to fuse the multi-scale feature, and use shortcut connection to incorporate the multi-level information for better noise removal. In addition, the channel attention mechanism is employed to make the network concentrate on the most relevant auxiliary information and features that are beneficial to the denoising process best. To ease the training procedure, we reconstruct the output through a residual mode rather than a straightforward prediction. Experimental results demonstrate that our proposed ADRN scheme outperforms the state-of-the-art methods both in quantitative and visual evaluations.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑