Deep neural networks (DNNs) have become increasingly popular in recent years. However, despite their many successes, DNNs may also err and produce incorrect and potentially fatal outputs in safety-critical settings, such as autonomous driving, medical diagnosis, and airborne collision avoidance systems. Much work has been put into detecting such erroneous behavior in DNNs, e.g., via testing or verification, but removing these errors after their detection has received lesser attention. We present here a new tool, called 3M-DNN, for repairing a given DNN, which is known to err on some set of inputs. The novel repair procedure implemented in 3M-DNN computes a modification to the network’s weights that corrects its behavior, and attempts to minimize this change via a sequence of calls to a backend, black-box DNN verification engine. To the best of our knowledge, our method is the first one that allows repairing the network by simultaneously modifying multiple layers. This is achieved by splitting the network into sub-networks, and applying a single-layer repairing technique to each component. We evaluated 3M-DNN tool on an extensive set of benchmarks, obtaining promising results.
|Original language||American English|
|Title of host publication||Software Verification and Formal Methods for ML-Enabled Autonomous Systems - 5th International Workshop, FoMLAS 2022, and 15th International Workshop, NSV 2022, Proceedings|
|Editors||Omri Isac, Guy Katz, Radoslav Ivanov, Nina Narodytska, Laura Nenzi|
|Publisher||Springer Science and Business Media Deutschland GmbH|
|Number of pages||21|
|State||Published - 2022|
|Event||5th International Workshop on Software Verification and Formal Methods for ML-Enables Autonomous Systems, FoMLAS 2022 and 15th International Workshop on Numerical Software Verification, NSV 2022 - Haifa, Israel|
Duration: 11 Aug 2022 → 11 Aug 2022
|Name||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Conference||5th International Workshop on Software Verification and Formal Methods for ML-Enables Autonomous Systems, FoMLAS 2022 and 15th International Workshop on Numerical Software Verification, NSV 2022|
|Period||11/08/22 → 11/08/22|
Bibliographical noteFunding Information:
Acknowledgement. This work was partially supported by the Israel Science Foundation (grant number 683/18) and the HUJI Federmann Cyber Security Center.
© 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.