Visual explanations from spiking neural networks using inter-spike intervals

Abstract By emulating biological features in brain, Spiking Neural Networks (SNNs) offer an energy-efficient alternative to conventional deep learning. To make SNNs ubiquitous, a ‘visual explanation’ technique for analysing and explaining the internal spike behavior of such temporal deep SNNs is cru...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Youngeun Kim, Priyadarshini Panda
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/21c9d821aa6d46349f250d5a2ab3e691
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:21c9d821aa6d46349f250d5a2ab3e691
record_format	dspace
spelling	oai:doaj.org-article:21c9d821aa6d46349f250d5a2ab3e6912021-12-02T18:13:45ZVisual explanations from spiking neural networks using inter-spike intervals10.1038/s41598-021-98448-02045-2322https://doaj.org/article/21c9d821aa6d46349f250d5a2ab3e6912021-09-01T00:00:00Zhttps://doi.org/10.1038/s41598-021-98448-0https://doaj.org/toc/2045-2322Abstract By emulating biological features in brain, Spiking Neural Networks (SNNs) offer an energy-efficient alternative to conventional deep learning. To make SNNs ubiquitous, a ‘visual explanation’ technique for analysing and explaining the internal spike behavior of such temporal deep SNNs is crucial. Explaining SNNs visually will make the network more transparent giving the end-user a tool to understand how SNNs make temporal predictions and why they make a certain decision. In this paper, we propose a bio-plausible visual explanation tool for SNNs, called Spike Activation Map (SAM). SAM yields a heatmap (i.e., localization map) corresponding to each time-step of input data by highlighting neurons with short inter-spike interval activity. Interestingly, without the use of gradients and ground truth, SAM produces a temporal localization map highlighting the region of interest in an image attributed to an SNN’s prediction at each time-step. Overall, SAM outsets the beginning of a new research area ‘explainable neuromorphic computing’ that will ultimately allow end-users to establish appropriate trust in predictions from SNNs.Youngeun KimPriyadarshini PandaNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 11, Iss 1, Pp 1-14 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Youngeun Kim Priyadarshini Panda Visual explanations from spiking neural networks using inter-spike intervals
description	Abstract By emulating biological features in brain, Spiking Neural Networks (SNNs) offer an energy-efficient alternative to conventional deep learning. To make SNNs ubiquitous, a ‘visual explanation’ technique for analysing and explaining the internal spike behavior of such temporal deep SNNs is crucial. Explaining SNNs visually will make the network more transparent giving the end-user a tool to understand how SNNs make temporal predictions and why they make a certain decision. In this paper, we propose a bio-plausible visual explanation tool for SNNs, called Spike Activation Map (SAM). SAM yields a heatmap (i.e., localization map) corresponding to each time-step of input data by highlighting neurons with short inter-spike interval activity. Interestingly, without the use of gradients and ground truth, SAM produces a temporal localization map highlighting the region of interest in an image attributed to an SNN’s prediction at each time-step. Overall, SAM outsets the beginning of a new research area ‘explainable neuromorphic computing’ that will ultimately allow end-users to establish appropriate trust in predictions from SNNs.
format	article
author	Youngeun Kim Priyadarshini Panda
author_facet	Youngeun Kim Priyadarshini Panda
author_sort	Youngeun Kim
title	Visual explanations from spiking neural networks using inter-spike intervals
title_short	Visual explanations from spiking neural networks using inter-spike intervals
title_full	Visual explanations from spiking neural networks using inter-spike intervals
title_fullStr	Visual explanations from spiking neural networks using inter-spike intervals
title_full_unstemmed	Visual explanations from spiking neural networks using inter-spike intervals
title_sort	visual explanations from spiking neural networks using inter-spike intervals
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/21c9d821aa6d46349f250d5a2ab3e691
work_keys_str_mv	AT youngeunkim visualexplanationsfromspikingneuralnetworksusinginterspikeintervals AT priyadarshinipanda visualexplanationsfromspikingneuralnetworksusinginterspikeintervals
_version_	1718378439325515776

Visual explanations from spiking neural networks using inter-spike intervals

Ejemplares similares