Title:	Neural Criticality Metric for Object Detection Deep Neural Networks
Authors:	Diviš, Václav Schuster, Tobias Hrúz, Marek
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
Issue Date:	2022
Publisher:	SPRINGER-VERLAG BERLIN
Document type:	konferenční příspěvek ConferenceObject
URI:	2-s2.0-85138995786 http://hdl.handle.net/11025/51464
ISBN:	978-3-031-14861-3
ISSN:	0302-9743
Keywords in different language:	DNN safety;Object detection;Neural criticality
Abstract in different language:	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.
Rights:	Plný text je přístupný v rámci univerzity přihlášeným uživatelům. © The Author(s), under exclusive licence to Springer Nature B.V.
Appears in Collections:	Konferenční příspěvky / Conference papers (NTIS) Konferenční příspěvky / Conference Papers (KKY) OBD

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.