Unsupervised Deep Non-rigid Alignment by Low-Rank Loss and Multi-input Attention
Unsupervised Deep Non-rigid Alignment by Low-Rank Loss and Multi-input Attention
复制标题
通过低秩损失和多输入注意力的无监督深度非刚性对齐
DOI：
10.1007/978-3-031-16446-0_18
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Bise Ryoma
中科院分区：
文献类型：
--
作者：
Asanomi Takanori;Nishimura Kazuya;Song Heon;Hayashida Junya;Sekiguchi Hiroyuki;Yagi Takayuki;Sato Imari;Bise Ryoma
我们提出了一种深度低阶配准网络，该网络可以同时对多幅图像进行非刚性配准和噪声分解，而不受严重的噪声和稀疏干扰。为了解决这一具有挑战性的任务，我们在深度学习中引入了低阶损失，假设一组排列良好、去噪良好的图像应该是线性相关的，因此，由图像组成的矩阵应该是低阶的。这允许我们以自监督学习方式(即，不需要监督数据)从输入图像中去除噪声和损坏。此外，我们还在暹罗U网中引入了多输入注意模块，以便从图像集合中聚合腐败信息。据我们所知，这是首次尝试将低阶损失引入基于深度学习的非刚性对齐。通过对合成数据和真实医学图像数据的实验，证明了该方法的有效性。该代码将在https://github.com/asanomitakanori/Unsupervised-Deep-Non-Rigid-Alignment-by-Low-Rank-Loss-and-Multi-Input-Attention.中公开提供
We propose a deep low-rank alignment network that can simultaneously perform non-rigid alignment and noise decomposition for multiple images despite severe noise and sparse corruptions. To address this challenging task, we introduce a low-rank loss in deep learning under the assumption that a set of well-aligned, well-denoised images should be linearly correlated, and thus, that a matrix consisting of the images should be low-rank. This allows us to remove the noise and corruption from input images in a self-supervised learning manner (i.e., without requiring supervised data). In addition, we introduce multi-input attention modules into Siamese U-nets in order to aggregate the corruption information from the set of images. To the best of our knowledge, this is the first attempt to introduce a low-rank loss for deep learning-based non-rigid alignment. Experiments using both synthetic data and real medical image data demonstrate the effectiveness of the proposed method. The code will be publicly available in https://github.com/asanomitakanori/Unsupervised-Deep-Non-Rigid-Alignment-by-Low-Rank-Loss-and-Multi-Input-Attention.