Distribution Awareness for AI System Testing

TL;DR

This paper introduces an out-of-distribution (OOD) guided testing method for deep learning systems that improves error detection relevance and enhances robustness by generating more meaningful unseen test cases.

Contribution

The paper proposes a novel OOD-guided testing approach that considers data distribution, improving error relevance assessment and robustness in DL system testing.

Findings

Filters up to 55.44% of error test cases on CIFAR-10

Increases robustness by 10.05%

Enhances test case relevance through distribution awareness

Abstract

As Deep Learning (DL) is continuously adopted in many safety critical applications, its quality and reliability start to raise concerns. Similar to the traditional software development process, testing the DL software to uncover its defects at an early stage is an effective way to reduce risks after deployment. Although recent progress has been made in designing novel testing techniques for DL software, the distribution of generated test data is not taken into consideration. It is therefore hard to judge whether the identified errors are indeed meaningful errors to the DL application. Therefore, we propose a new OOD-guided testing technique which aims to generate new unseen test cases relevant to the underlying DL system task. Our results show that this technique is able to filter up to 55.44% of error test case on CIFAR-10 and is 10.05% more effective in enhancing robustness.