Zur Kurzanzeige

dc.contributor.author
Rutishauser, Georg
dc.contributor.author
Scherer, Moritz
dc.contributor.author
Fischer, Tim
dc.contributor.author
Benini, Luca
dc.date.accessioned
2022-07-26T06:54:35Z
dc.date.available
2022-01-24T10:30:16Z
dc.date.available
2022-01-24T13:46:04Z
dc.date.available
2022-03-23T07:21:46Z
dc.date.available
2022-05-10T09:02:12Z
dc.date.available
2022-07-25T17:33:04Z
dc.date.available
2022-07-26T06:54:35Z
dc.date.issued
2022
dc.identifier.isbn
978-3-9819263-6-1
en_US
dc.identifier.isbn
978-1-6654-9637-7
en_US
dc.identifier.other
10.23919/DATE54114.2022.9774592
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/527816
dc.identifier.doi
10.3929/ethz-b-000527816
dc.description.abstract
Dynamic Vision Sensors (DVS) offer the opportunity to scale the energy consumption in image acquisition proportionally to the activity in the captured scene by only transmitting data when the captured image changes. Their potential for energy-proportional sensing makes them highly attractive for severely energy-constrained sensing nodes at the edge. Most approaches to the processing of DVS data employ Spiking Neural Networks to classify the input from the sensor. In this paper, we propose an alternative, event frame-based approach to the classification of DVS video data. We assemble ternary video frames from the event stream and process them with a fully ternarized Temporal Convolutional Network which can be mapped to CUTIE, a highly energy-efficient Ternary Neural Network accelerator. The network mapped to the accelerator achieves a classification accuracy of 94.5%, matching the state of the art for embedded implementations. We implement the processing pipeline in a modern 22 nm FDX technology and perform post-synthesis power simulation of the network running on the system, achieving an inference energy of 1.7 μJ which is 647x lower than previously reported results based on Spiking Neural Networks.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
IEEE
en_US
dc.rights.uri
http://creativecommons.org/licenses/by-nc-nd/4.0/
dc.subject
Dynamic vision sensor (DVS)
en_US
dc.subject
Computer Vision
en_US
dc.subject
Neural network
en_US
dc.title
Ternarized TCN for μJ/Inference Gesture Recognition from DVS Event Frames
en_US
dc.type
Conference Paper
dc.rights.license
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
dc.date.published
2022-05-19
ethz.book.title
2022 Design, Automation & Test in Europe Conference & Exhibition (DATE)
en_US
ethz.pages.start
736
en_US
ethz.pages.end
741
en_US
ethz.size
6 p. accepted version
en_US
ethz.version.deposit
acceptedVersion
en_US
ethz.event
25th Design, Automation and Test in Europe Conference (DATE 2022)
en_US
ethz.event.location
Online
en_US
ethz.event.date
March 14-23, 2022
en_US
ethz.notes
Conference lecture held on 22 March 2022.
en_US
ethz.grant
A Cognitive Fractal and Secure EDGE based on an unique Open-Safe-Reliable-Low Power Hardware Platform Node
en_US
ethz.grant
Heterogeneous Computing Systems with Customized Accelerators
en_US
ethz.identifier.wos
ethz.identifier.scopus
ethz.publication.place
Piscataway, NJ
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02636 - Institut für Integrierte Systeme / Integrated Systems Laboratory::03996 - Benini, Luca / Benini, Luca
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02636 - Institut für Integrierte Systeme / Integrated Systems Laboratory::03996 - Benini, Luca / Benini, Luca
en_US
ethz.grant.agreementno
877056
ethz.grant.agreementno
180625
ethz.grant.agreementno
877056
ethz.grant.agreementno
180625
ethz.grant.agreementno
877056
ethz.grant.agreementno
180625
ethz.grant.fundername
EC
ethz.grant.fundername
SNF
ethz.grant.fundername
EC
ethz.grant.fundername
SNF
ethz.grant.fundername
EC
ethz.grant.fundername
SNF
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.funderDoi
10.13039/501100001711
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.funderDoi
10.13039/501100001711
ethz.grant.funderDoi
10.13039/501100000780
ethz.grant.funderDoi
10.13039/501100001711
ethz.grant.program
H2020
ethz.grant.program
Croatian-Swiss Research Programme (CSRP)
ethz.grant.program
H2020
ethz.grant.program
Croatian-Swiss Research Programme (CSRP)
ethz.grant.program
H2020
ethz.grant.program
Croatian-Swiss Research Programme (CSRP)
ethz.date.deposited
2022-01-24T10:30:23Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2022-05-10T09:02:18Z
ethz.rosetta.lastUpdated
2023-02-07T04:49:32Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Ternarized%20TCN%20for%20%CE%BCJ/Inference%20Gesture%20Recognition%20from%20DVS%20Event%20Frames&rft.date=2022&rft.spage=736&rft.epage=741&rft.au=Rutishauser,%20Georg&Scherer,%20Moritz&Fischer,%20Tim&Benini,%20Luca&rft.isbn=978-3-9819263-6-1&978-1-6654-9637-7&rft.genre=proceeding&rft_id=info:doi/10.23919/DATE54114.2022.9774592&rft.btitle=2022%20Design,%20Automation%20&%20Test%20in%20Europe%20Conference%20&%20Exhibition%20(DATE)
 Printexemplar via ETH-Bibliothek suchen

Dateien zu diesem Eintrag

Thumbnail

Publikationstyp

Zur Kurzanzeige