Design Methodology for Deep Out-of-Distribution Detectors in Real-Time Cyber-Physical Systems

TL;DR

This paper presents a design methodology using genetic algorithms to optimize deep out-of-distribution detectors for real-time cyber-physical systems, balancing accuracy and response time for resource-constrained environments.

Contribution

It introduces a systematic approach to tune deep OOD detectors for embedded systems, including preprocessing optimization and task graph selection under ROS.

Findings

Significant reduction in response time while maintaining accuracy.

Effective optimization of preprocessing pipeline and quantization methods.

Demonstrated on two embedded platforms with variational autoencoder detectors.

Abstract

When machine learning (ML) models are supplied with data outside their training distribution, they are more likely to make inaccurate predictions; in a cyber-physical system (CPS), this could lead to catastrophic system failure. To mitigate this risk, an out-of-distribution (OOD) detector can run in parallel with an ML model and flag inputs that could lead to undesirable outcomes. Although OOD detectors have been well studied in terms of accuracy, there has been less focus on deployment to resource constrained CPSs. In this study, a design methodology is proposed to tune deep OOD detectors to meet the accuracy and response time requirements of embedded applications. The methodology uses genetic algorithms to optimize the detector's preprocessing pipeline and selects a quantization method that balances robustness and response time. It also identifies several candidate task graphs under the Robot Operating System (ROS) for deployment of the selected design. The methodology is demonstrated on two variational autoencoder based OOD detectors from the literature on two embedded platforms. Insights into the trade-offs that occur during the design process are provided, and it is shown that this design methodology can lead to a drastic reduction in response time in relation to an unoptimized OOD detector while maintaining comparable accuracy.