Title: | Neural Criticality: Validation of Convolutional Neural Networks |
Authors: | Diviš, Václav Hrúz, Marek |
Citation: | DIVIŠ, V. HRÚZ, M. Neural Criticality: Validation of Convolutional Neural Networks. In CEUR Workshop Proceedings. Neuveden: CEUR-WS, 2021. s. nestránkováno. ISBN: neuvedeno , ISSN: 1613-0073 |
Issue Date: | 2021 |
Publisher: | CEUR-WS |
Document type: | konferenční příspěvek ConferenceObject |
URI: | 2-s2.0-85101278318 http://hdl.handle.net/11025/47141 |
ISBN: | neuvedeno |
ISSN: | 1613-0073 |
Keywords in different language: | Criticality;Validation;Convolutional Neural Networks |
Abstract in different language: | The black-box behavior of Convolutional Neural Networks is one of the biggest obstacles to the development of a standardized validation process. Methods for analyzing and validating neural networks currently rely on approaches and metrics provided by the scientific community without considering functional safety requirements. However, automotive norms, such as ISO26262 and ISO/PAS21448, do require a comprehensive knowledge of the system and of the working environment in which the network will be deployed. In order to gain such a knowledge and mitigate the natural uncertainty of probabilistic models, we focused on investigating the influence of filter weights on the classification confidence in Single Point Of Failure fashion. We laid the theoretical foundation of a method called the Neurons’ Criticality Analysis. This method, as described in this article, helps evaluate the criticality of the tested network and choose related plausibility mechanism. Copyright © 2021, for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). |
Rights: | Plný text je přístupný v rámci univerzity přihlášeným uživatelům. © authors |
Appears in Collections: | Konferenční příspěvky / Conference Papers (KKY) OBD |
Files in This Item:
File | Size | Format | |
---|---|---|---|
Divis_Neural_Criticality_SafeAI_2021.pdf | 485,57 kB | Adobe PDF | View/Open Request a copy |
Please use this identifier to cite or link to this item:
http://hdl.handle.net/11025/47141
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.