Publication details
Runtime Monitoring for Out-of-Distribution Detection in Object Detection Neural Networks
| Authors | |
|---|---|
| Year of publication | 2023 |
| Type | Paper in proceedings |
| Conference | FORMAL METHODS, FM 2023 |
| MU Faculty or unit | |
| Citation | |
| Doi | https://doi.org/10.1007/978-3-031-27481-7_36 |
| Keywords | Runtime monitoring; Neural networks; Out-of-distribution detection; Object detection |
| Description | Runtime monitoring provides a more realistic and applicable alternative to verification in the setting of real neural networks used in industry. It is particularly useful for detecting out-of-distribution (OOD) inputs, for which the network was not trained and can yield erroneous results. We extend a runtime-monitoring approach previously proposed for classification networks to perception systems capable of identification and localization of multiple objects. Furthermore, we analyze its adequacy experimentally on different kinds of OOD settings, documenting the overall efficacy of our approach. |
| Related projects: |