标题： 训练样本的光谱预调制提高了相位提取神经网络 (PhENN) 的空间分辨率 显示英文标题

标题： Spectral pre-modulation of training examples enhances the spatial resolution of the Phase Extraction Neural Network (PhENN)

摘要： 基于深度机器学习的相位提取神经网络（PhENN）是一种计算架构，可以从原始强度数据中实现无透镜定量相位恢复。 PhENN 是一种深度卷积神经网络，通过由真实的相位物体及其对应的强度衍射图样组成的样本对进行训练；之后，给定一个测试的原始强度图样，PhENN 能够稳健地重建原始相位物体，在许多情况下，即使对于训练样本数据库之外的对象也能做到这一点。 在这里，我们展示了训练样本的空间频率内容是限制 PhENN 空间频率响应的重要因素。 例如，如果训练数据库在高空间频率上相对稀疏（大多数自然场景也是如此），那么 PhENN 在测试模式中解析精细空间特征的能力也会相应受限。 为了解决这个问题，我们建议在将训练样本呈现给 PhENN 之前“展平”其功率谱密度。 对于遵循自然场景统计的相位物体，我们通过实验表明，频谱预调制方法可以将 PhENN 的空间分辨率提高一倍。 显示英文摘要

摘要： The Phase Extraction Neural Network (PhENN) is a computational architecture, based on deep machine learning, for lens-less quantitative phase retrieval from raw intensity data. PhENN is a deep convolutional neural network trained through examples consisting of pairs of true phase objects and their corresponding intensity diffraction patterns; thereafter, given a test raw intensity pattern PhENN is capable of reconstructing the original phase object robustly, in many cases even for objects outside the database where the training examples were drawn from. Here, we show that the spatial frequency content of the training examples is an important factor limiting PhENN's spatial frequency response. For example, if the training database is relatively sparse in high spatial frequencies, as most natural scenes are, PhENN's ability to resolve fine spatial features in test patterns will be correspondingly limited. To combat this issue, we propose "flattening" the power spectral density of the training examples before presenting them to PhENN. For phase objects following the statistics of natural scenes, we demonstrate experimentally that the spectral pre-modulation method enhances the spatial resolution of PhENN by a factor of 2.