Adaptive Activation Thresholding: Dynamic Routing Type Behavior for Interpretability in Convolutional Neural Networks
Adaptive Activation Thresholding: Dynamic Routing Type Behavior for Interpretability in Convolutional Neural Networks
复制标题
DOI：
10.1109/iccv.2019.00504
复制
发表时间：
2019-10
期刊：
影响因子：
--
通讯作者：
Yiyou Sun;Sathya Ravi;Vikas Singh
中科院分区：
文献类型：
--
作者：
Yiyou Sun;Sathya Ravi;Vikas Singh
人们对可以帮助我们理解或解释神经网络的策略越来越感兴趣，也就是说，不仅提供预测，而且还提供额外的背景来解释为什么和如何。虽然许多当前的方法提供了工具来对给定的（训练的）网络事后进行这种分析，但最近的结果（特别是在胶囊网络上）表明，当类映射到网络前几层中的一些高级“概念”时，网络的行为更容易解释或解释。这种训练可以通过动态/EM路由来完成，其中用于各个类（或图像子集）的网络“路由”是动态的，并且即使整个网络可能不是稀疏的，也涉及很少的节点。在本文中，我们展示了SGD方案的简单修改如何有助于在卷积神经网络中提供动态/EM路由类型行为。通过大量的实验，我们评估了这种想法的效果，我们获得了有希望的结果，同时也表明，在可达到的准确性没有妥协。此外，我们表明，微小的修改似乎是ad-hoc，新算法可以分析的近似方法，可证明匹配已知的SGD率。
There is a growing interest in strategies that can help us understand or interpret neural networks -- that is, not merely provide a prediction, but also offer additional context explaining why and how. While many current methods offer tools to perform this analysis for a given (trained) network post-hoc, recent results (especially on capsule networks) suggest that when classes map to a few high level ``concepts'' in the preceding layers of the network, the behavior of the network is easier to interpret or explain. Such training may be accomplished via dynamic/EM routing where the network ``routes'' for individual classes (or subsets of images) are dynamic and involve few nodes even if the full network may not be sparse. In this paper, we show how a simple modification of the SGD scheme can help provide dynamic/EM routing type behavior in convolutional neural networks. Through extensive experiments, we evaluate the effect of this idea for interpretability where we obtain promising results, while also showing that no compromise in attainable accuracy is involved. Further, we show that the minor modification is seemingly ad-hoc, the new algorithm can be analyzed by an approximate method which provably matches known rates for SGD.