Topological Uncertainty: Monitoring trained neural networks through persistence of activation graphs
This addresses the need for post-training monitoring of neural networks in industrial applications where data may differ from training benchmarks, offering a novel approach without requiring architectural changes or retraining.
The authors tackled the problem of monitoring trained neural networks for reliability in open-world settings by developing a method based on topological properties of activation graphs, which achieved effective performance in network selection, out-of-distribution detection, and shift-detection across various datasets.
Although neural networks are capable of reaching astonishing performances on a wide variety of contexts, properly training networks on complicated tasks requires expertise and can be expensive from a computational perspective. In industrial applications, data coming from an open-world setting might widely differ from the benchmark datasets on which a network was trained. Being able to monitor the presence of such variations without retraining the network is of crucial importance. In this article, we develop a method to monitor trained neural networks based on the topological properties of their activation graphs. To each new observation, we assign a Topological Uncertainty, a score that aims to assess the reliability of the predictions by investigating the whole network instead of its final layer only, as typically done by practitioners. Our approach entirely works at a post-training level and does not require any assumption on the network architecture, optimization scheme, nor the use of data augmentation or auxiliary datasets; and can be faithfully applied on a large range of network architectures and data types. We showcase experimentally the potential of Topological Uncertainty in the context of trained network selection, Out-Of-Distribution detection, and shift-detection, both on synthetic and real datasets of images and graphs.