# Towards Improved Testing For Deep Learning

## TL;DR

This paper reviews current deep neural network testing methods, highlights their limitations, and proposes a new coverage criterion aimed at improving the effectiveness and scalability of testing in safety-critical applications.

## Contribution

It introduces a novel coverage criterion designed specifically for deep neural networks to enhance testing comprehensiveness and scalability.

## Key findings

- Existing testing methods are insufficient for safety-critical DNN applications.
- The proposed coverage criterion better captures the internal logic of DNNs.
- Scalable and generalizable testing approaches are necessary for deployment in real-world scenarios.

## Abstract

The growing use of deep neural networks in safety-critical applications makes it necessary to carry out adequate testing to detect and correct any incorrect behavior for corner case inputs before they can be actually used. Deep neural networks lack an explicit control-flow structure, making it impossible to apply to them traditional software testing criteria such as code coverage. In this paper, we examine existing testing methods for deep neural networks, the opportunities for improvement and the need for a fast, scalable, generalizable end-to-end testing method. We also propose a coverage criterion for deep neural networks that tries to capture all possible parts of the deep neural network's logic.