Reducing the U-Net size for practical scenarios: Virus recognition in electron microscopy images
Reducing the U-Net size for practical scenarios: Virus recognition in electron microscopy images
复制标题
DOI：
10.1016/j.cmpb.2019.05.026
复制
发表时间：
2019-09-01
影响因子：
6.1
通讯作者：
Sintorn, Ida-Maria
中科院分区：
文献类型：
--
作者：
Matuszewski, Damian J.;Sintorn, Ida-Maria
背景和目的：卷积神经网络（cnn）提供了类似人类专家的性能，同时它们的预测速度更快，更一致。然而，大多数提议的cnn都需要昂贵的最先进的硬件，这极大地限制了它们在实际场景和商业系统中的使用，特别是在临床、生物医学和其他需要实时分析的应用中。在本文中，我们通过对一个流行的CNN: U-Net的结构进行参数化和减少可训练权值的数量来研究使CNN更轻化的可能性。方法：为了证明可比的结果可以用比原始U-Net少得多的可训练权重来实现，我们在透射电子显微镜图像中使用了具有挑战性的逐像素病毒分类应用，注释最少（即仅由病毒颗粒中心或中心线组成）。我们研究了4个U-Net超参数：基本特征映射的数量，特征映射的乘数，编解码层的数量和最后2个卷积层的特征映射的数量。结果：我们的实验得出了两个主要结论：1)如果要使用较少的可训练权值，那么架构超参数是关键的；2)如果对可训练权值没有限制，那么使用更深的网络通常会得到更好的结果。然而，训练更大的网络需要更长的时间，通常需要更多的数据，这样的网络也更容易过度拟合。我们的最佳模型达到了82.2%的准确率，与原始的U-Net相似，同时使用了近4倍的可训练权值（7.8 M与31.0 M相比）。我们还提出了一个具有< 2M可训练权值的网络，其准确率达到76.4%。结论：提出的U-Net超参数探索方法可适用于其他cnn和其他应用。它允许以更有效的可训练权重使用为目标进行全面的CNN架构设计。使网络更快、更轻对于它们在许多实际应用中的实现至关重要。此外，一个较轻的网络应该不太容易过度拟合，因此泛化更好。(C) 2019年Elsevier B.V.出版
Background and objective: Convolutional neural networks (CNNs) offer human experts-like performance and in the same time they are faster and more consistent in their prediction. However, most of the proposed CNNs require an expensive state-of-the-art hardware which substantially limits their use in practical scenarios and commercial systems, especially for clinical, biomedical and other applications that require on-the-fly analysis. In this paper, we investigate the possibility of making CNNs lighter by parametrizing the architecture and decreasing the number of trainable weights of a popular CNN: U-Net.Methods: In order to demonstrate that comparable results can be achieved with substantially less trainable weights than the original U-Net we used a challenging application of a pixel-wise virus classification in Transmission Electron Microscopy images with minimal annotations (i.e. consisting only of the virus particle centers or centerlines). We explored 4 U-Net hyper-parameters: the number of base feature maps, the feature maps multiplier, the number of the encoding-decoding levels and the number of feature maps in the last 2 convolutional layers.Results: Our experiments lead to two main conclusions: 1) the architecture hyper-parameters are pivotal if less trainable weights are to be used, and 2) if there is no restriction on the trainable weights number using a deeper network generally gives better results. However, training larger networks takes longer, typically requires more data and such networks are also more prone to overfitting. Our best model achieved an accuracy of 82.2% which is similar to the original U-Net while using nearly 4 times less trainable weights (7.8 M in comparison to 31.0 M). We also present a network with < 2M trainable weights that achieved an accuracy of 76.4%.Conclusions: The proposed U-Net hyper-parameter exploration can be adapted to other CNNs and other applications. It allows a comprehensive CNN architecture designing with the aim of a more efficient trainable weight use. Making the networks faster and lighter is crucial for their implementation in many practical applications. In addition, a lighter network ought to be less prone to over-fitting and hence generalize better. (C) 2019 Published by Elsevier B.V.