Runtime Monitoring DNN-Based Perception

TL;DR

This paper reviews runtime monitoring techniques for DNN-based perception systems, emphasizing the importance of rigorous monitor design to ensure safety-critical applications operate reliably beyond static verification.

Contribution

It provides an overview of classical and formal methods for runtime verification of DNN perception, highlighting differences in monitor design and the importance of data outside the operational domain.

Findings

Classical and formal methods for runtime monitoring are discussed.

Design of monitors varies between machine learning and formal methods communities.

Data outside the operational domain is crucial for monitor effectiveness.

Abstract

Deep neural networks (DNNs) are instrumental in realizing complex perception systems. As many of these applications are safety-critical by design, engineering rigor is required to ensure that the functional insufficiency of the DNN-based perception is not the source of harm. In addition to conventional static verification and testing techniques employed during the design phase, there is a need for runtime verification techniques that can detect critical events, diagnose issues, and even enforce requirements. This tutorial aims to provide readers with a glimpse of techniques proposed in the literature. We start with classical methods proposed in the machine learning community, then highlight a few techniques proposed by the formal methods community. While we surely can observe similarities in the design of monitors, how the decision boundaries are created vary between the two communities. We conclude by highlighting the need to rigorously design monitors, where data availability outside the operational domain plays an important role.