Convolutional Neural Networks (CNNs) currently achieve state-of-the-art accuracy in image classification. With a growing number of classes, the accuracy usually drops as the possibilities of confusion increase. Interestingly, the class confusion patterns follow a hierarchical structure over the classes. We present visual-analytics methods to reveal and analyze this hierarchy of similar classes in relation with CNN-internal data. We found that this hierarchy not only dictates the confusion patterns between the classes, it furthermore dictates the learning behavior of CNNs. In particular, the early layers in these networks develop feature detectors that can separate high-level groups of classes quite well, even after a few training epochs. In contrast, the latter layers require substantially more epochs to develop specialized feature detectors that can separate individual classes. We demonstrate how these insights are key to significant improvement in accuracy by designing hierarchy-aware CNNs that accelerate model convergence and alleviate overfitting. We further demonstrate how our methods help in identifying various quality issues in the training data.

本文研究了卷积神经网络的分类问题，并发现不同类别之间存在层级关系。同时，网络中不同层次的特征检测器的训练行为也在一定程度上受到类别层次结构的影响。根据这些发现，设计了一种层次感知的卷积神经网络，有效提高了模型的准确性并解决过拟合问题，并能够帮助发现训练数据中的各种质量问题。

卷积神经网络是否学习类别层次结构？