Datamorphic Testing: A Methodology for Testing AI Applications
Abstract
With the rapid growth of the applications of machine learning (ML) and other artificial intelligence (AI) techniques, adequate testing has become a necessity to ensure their quality. This paper identifies the characteristics of AI applications that distinguish them from traditional software, and analyses the main difficulties in applying existing testing methods. Based on this analysis, we propose a new method called datamorphic testing and illustrate the method with an example of testing face recognition applications. We also report an experiment with four real industrial application systems of face recognition to validate the proposed approach.
Cite
@article{arxiv.1912.04900,
title = {Datamorphic Testing: A Methodology for Testing AI Applications},
author = {Hong Zhu and Dongmei Liu and Ian Bayley and Rachel Harrison and Fabio Cuzzolin},
journal= {arXiv preprint arXiv:1912.04900},
year = {2019}
}
Comments
This technical report is an extended version of conference paper: [Zhu, H., Liu, D., Ian Bayley, I., Harrison, R. and Cuzzolin, F., Datamorphic Testing: A Method for Testing Intelligent Applications, The 1st IEEE International Conference On Artificial Intelligence Testing (IEEE AITest 2019), San Francisco, California, USA, April, 4 - 9, 2019.]