TY - GEN
T1 - HUDD
T2 - 44th ACM/IEEE International Conference on Software Engineering: Companion, ICSE-Companion 2022
AU - Fahmy, Hazem
AU - Pastore, Fabrizio
AU - Briand, Lionel
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - We present HUDD, a tool that supports safety analysis practices for systems enabled by Deep Neural Networks (DNNs) by automatically identifying the root causes for DNN errors and retraining the DNN. HUDD stands for Heatmap-based Unsupervised Debugging of DNNs, it automatically clusters error-inducing images whose results are due to common subsets of DNN neurons. The intent is for the generated clusters to group error-inducing images having common characteristics, that is, having a common root cause.HUDD identifies root causes by applying a clustering algorithm to matrices (i.e., heatmaps) capturing the relevance of every DNN neuron on the DNN outcome. Also, HUDD retrains DNNs with images that are automatically selected based on their relatedness to the identified image clusters. Our empirical evaluation with DNNs from the automotive domain have shown that HUDD automatically identifies all the distinct root causes of DNN errors, thus supporting safety analysis. Also, our retraining approach has shown to be more effective at improving DNN accuracy than existing approaches. A demo video of HUDD is available at https://youtu.be/drjVakP7jdU.
AB - We present HUDD, a tool that supports safety analysis practices for systems enabled by Deep Neural Networks (DNNs) by automatically identifying the root causes for DNN errors and retraining the DNN. HUDD stands for Heatmap-based Unsupervised Debugging of DNNs, it automatically clusters error-inducing images whose results are due to common subsets of DNN neurons. The intent is for the generated clusters to group error-inducing images having common characteristics, that is, having a common root cause.HUDD identifies root causes by applying a clustering algorithm to matrices (i.e., heatmaps) capturing the relevance of every DNN neuron on the DNN outcome. Also, HUDD retrains DNNs with images that are automatically selected based on their relatedness to the identified image clusters. Our empirical evaluation with DNNs from the automotive domain have shown that HUDD automatically identifies all the distinct root causes of DNN errors, thus supporting safety analysis. Also, our retraining approach has shown to be more effective at improving DNN accuracy than existing approaches. A demo video of HUDD is available at https://youtu.be/drjVakP7jdU.
KW - DNN debugging
KW - Functional Safety Analysis
UR - http://www.scopus.com/inward/record.url?scp=85132439274&partnerID=8YFLogxK
U2 - 10.1109/ICSE-Companion55297.2022.9793750
DO - 10.1109/ICSE-Companion55297.2022.9793750
M3 - Conference contribution
AN - SCOPUS:85132439274
T3 - Proceedings - International Conference on Software Engineering
SP - 100
EP - 104
BT - Proceedings - 2022 ACM/IEEE 44th International Conference on Software Engineering
PB - IEEE Computer Society
Y2 - 22 May 2022 through 27 May 2022
ER -