Repairing Deep Neural Networks: Fix Patterns and Challenges

By: Md Johirul Islam, Rangeet Pan, Giang Nguyen, and Hridesh Rajan

Abstract

Significant interest in applying Deep Neural Network (DNN) has fueled the need to support engineering of software that uses DNNs. Repairing software that uses DNNs is one such unmistakable SE need where automated tools could be very helpful; however, we do not fully understand challenges to repairing and patterns that are utilized when manually repairing them. What challenges should automated repair tools address? What are the repair patterns whose automation could help developers? Which repair patterns should be assigned a higher priority for automation? This work presents a comprehensive study of bug fix patterns to address these questions. We have studied 415 repairs from Stack Overflow and 555 repairs from GitHub for five popular deep learning libraries Caffe, Keras, Tensorflow, Theano, and Torch to understand challenges in repairs and bug repair patterns. Our key findings reveal that DNN bug fix patterns are distinctive compared to traditional bug fix patterns; the most common bug fix patterns are fixing data dimension and neural network connectivity; DNN bug fixes have the potential to introduce adversarial vulnerabilities; DNN bug fixes frequently introduce new bugs; and DNN bug localization, reuse of trained model, and coping with frequent releases are major challenges faced by developers when fixing bugs. We also contribute a benchmark of 667 DNN (bug, repair) instances.

ACM Reference

Islam, M.J. et al. 2020. Repairing Deep Neural Networks: Fix Patterns and Challenges. ICSE: 42nd International Conference on Software Engineering, Seoul, South Korea (2020), 1135–1146.

BibTeX Reference

@inproceedings{IslamPanNguyenRajan2020,
  author = {Md Johirul Islam and Rangeet Pan and Giang Nguyen and Hridesh Rajan},
  title = {Repairing Deep Neural Networks: Fix Patterns and Challenges},
  booktitle = {ICSE: 42nd International Conference on Software Engineering, Seoul, South Korea},
  pages = {1135--1146},
  year = {2020},
  publisher = {{ACM}},
  editor = {Gregg Rothermel and Doo{-}Hwan Bae},
  doi = {10.1145/3377811.3380378},
  abstract = {
  Significant interest in applying Deep Neural Network (DNN)
  has fueled the need to support engineering of software that
  uses DNNs.
  Repairing software that uses DNNs is one such unmistakable SE
  need where automated tools could be very helpful; however,
  we do not fully understand challenges to repairing and
  patterns that are utilized when manually repairing them.
  What challenges should automated repair tools address?
  What are the repair patterns whose automation could help
  developers? Which repair patterns should be assigned a
  higher priority for automation?
  This work presents a comprehensive study of bug fix patterns
  to address these questions.
  We have studied 415 repairs from Stack Overflow and 555 repairs from
  GitHub for five popular deep learning libraries Caffe, Keras,
  Tensorflow, Theano, and Torch to understand challenges in
  repairs and bug repair patterns.
  Our key findings reveal that
  DNN bug fix patterns are distinctive compared to traditional bug fix patterns;  
  the most common bug fix patterns are fixing data dimension and neural network connectivity;
  DNN bug fixes have the potential to introduce adversarial vulnerabilities;
  DNN bug fixes frequently introduce new bugs; and
  DNN bug localization, reuse of trained model, and coping with frequent releases
  are major challenges faced by developers when fixing bugs.
  We also contribute a benchmark of 667 DNN (bug, repair) instances.},
}