An Overview of Structural Coverage Metrics for Testing Neural Networks

TL;DR

This paper reviews various structural coverage metrics for testing deep neural networks, evaluates them on perception and autonomous models, and introduces a tool, DNNCov, for measuring and reporting coverage to improve testing practices.

Contribution

It provides a comprehensive overview of existing coverage metrics, evaluates their effectiveness on real-world models, and introduces DNNCov, a tool for measuring and comparing coverage measures.

Findings

Coverage metrics vary in effectiveness across models.

DNNCov facilitates comprehensive coverage assessment.

The overview aids in selecting appropriate testing metrics.

Abstract

Deep neural network (DNN) models, including those used in safety-critical domains, need to be thoroughly tested to ensure that they can reliably perform well in different scenarios. In this article, we provide an overview of structural coverage metrics for testing DNN models, including neuron coverage (NC), k-multisection neuron coverage (kMNC), top-k neuron coverage (TKNC), neuron boundary coverage (NBC), strong neuron activation coverage (SNAC) and modified condition/decision coverage (MC/DC). We evaluate the metrics on realistic DNN models used for perception tasks (including LeNet-1, LeNet-4, LeNet-5, and ResNet20) as well as on networks used in autonomy (TaxiNet). We also provide a tool, DNNCov, which can measure the testing coverage for all these metrics. DNNCov outputs an informative coverage report to enable researchers and practitioners to assess the adequacy of DNN testing, compare different coverage measures, and to more conveniently inspect the model's internals during testing.