Cats Are Not Fish: Deep Learning Testing Calls for Out-Of-Distribution Awareness

As Deep Learning (DL) is continuously adopted in many industrial 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 depl...

Full description

Saved in:
Bibliographic Details
Published in:2020 35th IEEE/ACM International Conference on Automated Software Engineering (ASE) pp. 1041 - 1052
Main Authors: Berend, David, Xie, Xiaofei, Ma, Lei, Zhou, Lingjun, Liu, Yang, Xu, Chi, Zhao, Jianjun
Format: Conference Proceeding
Language:English
Published: ACM 01.09.2020
Subjects:
Data models
Deep learning
Deep learning testing
out of distribution
quality assurance
Software
Software engineering
Software reliability
Testing
Training data
ISSN:2643-1572
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As Deep Learning (DL) is continuously adopted in many industrial 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. According to the fundamental assumption of deep learning, the DL software does not provide statistical guarantee and has limited capability in handling data that falls outside of its learned distribution, i.e., out-of-distribution (OOD) data. Although recent progress has been made in designing novel testing techniques for DL software, which can detect thousands of errors, the current state-of-the-art DL testing techniques usually do not take the distribution of generated test data into consideration. It is therefore hard to judge whether the "identified errors" are indeed meaningful errors to the DL application (i.e., due to quality issues of the model) or outliers that cannot be handled by the current model (i.e., due to the lack of training data). Tofill this gap, we take thefi rst step and conduct a large scale empirical study, with a total of 451 experiment configurations, 42 deep neural networks (DNNs) and 1.2 million test data instances, to investigate and characterize the impact of OOD-awareness on DL testing. We further analyze the consequences when DL systems go into production by evaluating the effectiveness of adversarial retraining with distribution-aware errors. The results confirm that introducing data distribution awareness in both testing and enhancement phases outperforms distribution unaware retraining by up to 21.5%.
ISSN:2643-1572
DOI:10.1145/3324884.3416609