Neural Criticality Metric for Object Detection Deep Neural Networks
| dc.contributor.author | Diviš, Václav | |
| dc.contributor.author | Schuster, Tobias | |
| dc.contributor.author | Hrúz, Marek | |
| dc.date.accessioned | 2023-02-13T11:00:21Z | |
| dc.date.available | 2023-02-13T11:00:21Z | |
| dc.date.issued | 2022 | |
| dc.description.abstract-translated | The complexity of state-of-the-art Deep Neural Network (DNN) architectures exacerbates the search for safety relevant metrics and methods that could be used for functional safety assessments. In this article, we investigate Neurons' Criticality (the ability to affect the decision process) for several object detection DNN architectures. As a first step, we introduce the Neural Criticality metric for object detection DNNs and set a theoretical background. Subsequently, by conducting experiments, we verify that removing one neuron from the computational graph of a DNN can have a significant (positive, as well as negative) influence on the prediction's precision (object classification and localization). Finally, we build statistics for each neuron from pre-trained networks on the COCO object detection validation dataset and examine the network stability for the most critical neurons in order to prove our metric's validity. | en |
| dc.format | 13 s. | cs |
| dc.format.mimetype | application/pdf | |
| dc.identifier.citation | DIVIŠ, V. SCHUSTER, T. HRÚZ, M. Neural Criticality Metric for Object Detection Deep Neural Networks. In Computer Safety, Reliability and Security, SAFECOMP 2022 Workshops. BERLIN: SPRINGER-VERLAG BERLIN, 2022. s. 276-288. ISBN: 978-3-031-14861-3 , ISSN: 0302-9743 | cs |
| dc.identifier.document-number | 866543800026 | |
| dc.identifier.doi | 10.1007/978-3-031-14862-0_20 | |
| dc.identifier.isbn | 978-3-031-14861-3 | |
| dc.identifier.issn | 0302-9743 | |
| dc.identifier.obd | 43937110 | |
| dc.identifier.uri | 2-s2.0-85138995786 | |
| dc.identifier.uri | http://hdl.handle.net/11025/51464 | |
| dc.language.iso | en | en |
| dc.project.ID | TN01000024/Národní centrum kompetence - Kybernetika a umělá inteligence | cs |
| dc.publisher | SPRINGER-VERLAG BERLIN | en |
| dc.relation.ispartofseries | Computer Safety, Reliability and Security, SAFECOMP 2022 Workshops | en |
| dc.rights | Plný text je přístupný v rámci univerzity přihlášeným uživatelům. | cs |
| dc.rights | © The Author(s), under exclusive licence to Springer Nature B.V. | en |
| dc.rights.access | restrictedAccess | en |
| dc.subject.translated | DNN safety | en |
| dc.subject.translated | Object detection | en |
| dc.subject.translated | Neural criticality | en |
| dc.title | Neural Criticality Metric for Object Detection Deep Neural Networks | en |
| dc.type | konferenční příspěvek | cs |
| dc.type | ConferenceObject | en |
| dc.type.status | Peer-reviewed | en |
| dc.type.version | publishedVersion | en |
Files
Original bundle
1 - 1 out of 1 results
No Thumbnail Available
- Name:
- Divis_Schuster_Hruz_Neural_Criticality_Metric_SAFECOMP_2022.pdf
- Size:
- 664.49 KB
- Format:
- Adobe Portable Document Format