PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices

Rahim Entezari; Olga Saukh

doi:10.1109/IPSN48710.2020.00067

PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices

Rahim Entezari, Olga Saukh

Institut für Technische Informatik (4480)

Publikation: Beitrag in Buch/Bericht/Konferenzband › Beitrag in einem Konferenzband › Begutachtung

Abstract

Today's deep neural networks require substantial computation resources for their training, storage, and inference, which limits their effective use on resource-constrained devices. Many recent research activities explore different options for compressing and optimizing deep models for the Internet of Things (IoT). Our work so far has focused on two important aspects of deep neural network compression: class-dependent model compression and explainable compression. We shortly summarize our contributions and conclude with an outline of our future research directions.

Originalsprache	englisch
Titel	Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020
Seiten	385-386
Seitenumfang	2
ISBN (elektronisch)	978-1-7281-5497-8
DOIs	https://doi.org/10.1109/IPSN48710.2020.00067
Publikationsstatus	Veröffentlicht - 21 Apr. 2020
Veranstaltung	19th ACM/IEEE International Conference on Information Processing in Sensor Networks: IPSN 2020 - Virtuell, Sydney, Australien Dauer: 21 Apr. 2020 → 24 Apr. 2020 Konferenznummer: 19 https://ipsn.acm.org/2020/

Publikationsreihe

Name	Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020

Konferenz

Konferenz	19th ACM/IEEE International Conference on Information Processing in Sensor Networks
Kurztitel	IPSN 2020
Land/Gebiet	Australien
Ort	Sydney
Zeitraum	21/04/20 → 24/04/20
Internetadresse	https://ipsn.acm.org/2020/

ASJC Scopus subject areas

Informationssysteme und -management
Signalverarbeitung
Computernetzwerke und -kommunikation

Zugriff auf Dokument

10.1109/IPSN48710.2020.00067

Andere Dateien und Links

http://www.scopus.com/inward/record.url?scp=85086895679&partnerID=8YFLogxK

Dieses zitieren

Entezari, R., & Saukh, O. (2020). PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices. in Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020 (S. 385-386). Artikel 9110987 (Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020). https://doi.org/10.1109/IPSN48710.2020.00067

PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices. / Entezari, Rahim; Saukh, Olga.
Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020. 2020. S. 385-386 9110987 (Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020).

Publikation: Beitrag in Buch/Bericht/Konferenzband › Beitrag in einem Konferenzband › Begutachtung

Entezari, R & Saukh, O 2020, PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices. in Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020., 9110987, Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020, S. 385-386, 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, Sydney, Australien, 21/04/20. https://doi.org/10.1109/IPSN48710.2020.00067

Entezari R, Saukh O. PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices. in Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020. 2020. S. 385-386. 9110987. (Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020). doi: 10.1109/IPSN48710.2020.00067

@inproceedings{368bb30f5de8452e8e345dbd04d82991,

title = "PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices",

abstract = "Today's deep neural networks require substantial computation resources for their training, storage, and inference, which limits their effective use on resource-constrained devices. Many recent research activities explore different options for compressing and optimizing deep models for the Internet of Things (IoT). Our work so far has focused on two important aspects of deep neural network compression: class-dependent model compression and explainable compression. We shortly summarize our contributions and conclude with an outline of our future research directions.",

keywords = "deep neural networks, Internet of Things, model compression",

author = "Rahim Entezari and Olga Saukh",

year = "2020",

month = apr,

day = "21",

doi = "10.1109/IPSN48710.2020.00067",

language = "English",

series = "Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020",

pages = "385--386",

booktitle = "Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020",

note = "19th ACM/IEEE International Conference on Information Processing in Sensor Networks : IPSN 2020, IPSN 2020 ; Conference date: 21-04-2020 Through 24-04-2020",

url = "https://ipsn.acm.org/2020/",

}

TY - GEN

T1 - PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices

AU - Entezari, Rahim

AU - Saukh, Olga

N1 - Conference code: 19

PY - 2020/4/21

Y1 - 2020/4/21

N2 - Today's deep neural networks require substantial computation resources for their training, storage, and inference, which limits their effective use on resource-constrained devices. Many recent research activities explore different options for compressing and optimizing deep models for the Internet of Things (IoT). Our work so far has focused on two important aspects of deep neural network compression: class-dependent model compression and explainable compression. We shortly summarize our contributions and conclude with an outline of our future research directions.

AB - Today's deep neural networks require substantial computation resources for their training, storage, and inference, which limits their effective use on resource-constrained devices. Many recent research activities explore different options for compressing and optimizing deep models for the Internet of Things (IoT). Our work so far has focused on two important aspects of deep neural network compression: class-dependent model compression and explainable compression. We shortly summarize our contributions and conclude with an outline of our future research directions.

KW - deep neural networks

KW - Internet of Things

KW - model compression

UR - http://www.scopus.com/inward/record.url?scp=85086895679&partnerID=8YFLogxK

U2 - 10.1109/IPSN48710.2020.00067

DO - 10.1109/IPSN48710.2020.00067

M3 - Conference paper

T3 - Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020

SP - 385

EP - 386

BT - Proceedings - 2020 19th ACM/IEEE International Conference on Information Processing in Sensor Networks, IPSN 2020

T2 - 19th ACM/IEEE International Conference on Information Processing in Sensor Networks

Y2 - 21 April 2020 through 24 April 2020

ER -

PhD Forum Abstract: Understanding Deep Model Compression for IoT Devices

Abstract

Publikationsreihe

Konferenz

ASJC Scopus subject areas

Zugriff auf Dokument

Andere Dateien und Links

Fingerprint

Dieses zitieren