In this paper, we propose a wide-field, high-resolution and photo-realistic microscope based on deep learning termed MASRM and design a deep residual neural network for reconstruction. We build an experimental system and a large-scale human blood image dataset. The experiments show that our model achieves better results and consumes much less time than conventional methods. The paper should be of interest to readers of your journal in the areas of microscopy and deep learning.

中文翻译：

---

使用多角度超分辨率的高通量深度学习显微镜


在本文中，我们提出了一种基于深度学习的宽视场、高分辨率和照片级真实感显微镜，称为 MASRM，并设计了用于重建的深度残差神经网络。我们建立了一个实验系统和一个大规模的人体血液图像数据集。实验表明，我们的模型比传统方法取得了更好的结果，并且消耗的时间少得多。显微镜和深度学习领域的期刊读者应该会对这篇论文感兴趣。