Activation Anomaly Analysis

Inspired by recent advances in coverage-guided analysis of neural networks, we propose a novel anomaly detection method. We show that the hidden activation values contain information useful to distinguish between normal and anomalous samples. Our approach combines three networks purely data-driven end-to-end model. Based on target network, alarm network decides if given sample is normal. Thanks our method even works strict semi-supervised settings. Strong results are achieved common data sets surpassing current baseline methods. allows inspect large amounts for anomalies across various applications.