论文标题
CNN神经元学习什么：可视化和聚类
What do CNN neurons learn: Visualization & Clustering
论文作者
论文摘要
近年来，卷积神经网络（CNN）在各种任务中表现出了惊人的进步。但是，尽管表现高，但训练和预测过程仍然是一个黑匣子，这是提取神经元在CNN中学习的东西是一个谜。在本文中，我们解决了从输入图像的焦点和偏好方面解释CNN的问题，以及神经元对具体最终预测的统治，激活和贡献。具体来说，我们使用两种技术 - 可视化和聚类 - 解决上述问题。可视化是指图像像素上梯度下降的方法，在聚类中，提出了两个算法分别以图像类别和网络神经元的群体聚类。实验和定量分析已经证明了这两种方法在解释问题：神经元学习什么。
In recent years convolutional neural networks (CNN) have shown striking progress in various tasks. However, despite the high performance, the training and prediction process remains to be a black box, leaving it a mystery to extract what neurons learn in CNN. In this paper, we address the problem of interpreting a CNN from the aspects of the input image's focus and preference, and the neurons' domination, activation and contribution to a concrete final prediction. Specifically, we use two techniques - visualization and clustering - to tackle the problems above. Visualization means the method of gradient descent on image pixel, and in clustering section two algorithms are proposed to cluster respectively over image categories and network neurons. Experiments and quantitative analyses have demonstrated the effectiveness of the two methods in explaining the question: what do neurons learn.
