Quannet: Joint image compression and classification over channels with limited bandwidth

Lahiru D. Chamain; Sen Ching Samson Cheung; Zhi Ding

doi:10.1109/ICME.2019.00066

Quannet: Joint image compression and classification over channels with limited bandwidth

Lahiru D. Chamain, Sen Ching Samson Cheung, Zhi Ding

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

18 Scopus citations

Abstract

The performance of cloud based image classification depends critically on its allocated bandwidth. Traditional data compression methods can negatively impact classification accuracy under limited bandwidth. We investigate the design of bandwidth efficient quantization for image encoding and compression with minimum classification accuracy loss. This work develops a simple neural network framework for joint quantization and classification. The proposed 'QuanNet' can optimize the quantization intervals of JPEG2000 encoder to minimize the classification loss. We show that our quantizer optimization can achieve significant accuracy improvement for a given channel bandwidth. Similarly, significant bandwidth can be saved to achieve a desired accuracy for cloud based image classification.

Original language	English
Title of host publication	Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019
Pages	338-343
Number of pages	6
ISBN (Electronic)	9781538695524
DOIs	https://doi.org/10.1109/ICME.2019.00066
State	Published - Jul 2019
Event	2019 IEEE International Conference on Multimedia and Expo, ICME 2019 - Shanghai, China Duration: Jul 8 2019 → Jul 12 2019

Publication series

Name	Proceedings - IEEE International Conference on Multimedia and Expo
Volume	2019-July
ISSN (Print)	1945-7871
ISSN (Electronic)	1945-788X

Conference

Conference	2019 IEEE International Conference on Multimedia and Expo, ICME 2019
Country/Territory	China
City	Shanghai
Period	7/8/19 → 7/12/19

Bibliographical note

Publisher Copyright:
© 2019 IEEE.

Keywords

Classification
Compression
Learning

ASJC Scopus subject areas

Computer Networks and Communications
Computer Science Applications

Access to Document

10.1109/ICME.2019.00066

Cite this

Chamain, L. D., Cheung, S. C. S., & Ding, Z. (2019). Quannet: Joint image compression and classification over channels with limited bandwidth. In Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019 (pp. 338-343). Article 8784889 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2019-July). https://doi.org/10.1109/ICME.2019.00066

@inproceedings{a8a138e10efa42a1a546bf7a7a9eebf8,

title = "Quannet: Joint image compression and classification over channels with limited bandwidth",

abstract = "The performance of cloud based image classification depends critically on its allocated bandwidth. Traditional data compression methods can negatively impact classification accuracy under limited bandwidth. We investigate the design of bandwidth efficient quantization for image encoding and compression with minimum classification accuracy loss. This work develops a simple neural network framework for joint quantization and classification. The proposed 'QuanNet' can optimize the quantization intervals of JPEG2000 encoder to minimize the classification loss. We show that our quantizer optimization can achieve significant accuracy improvement for a given channel bandwidth. Similarly, significant bandwidth can be saved to achieve a desired accuracy for cloud based image classification.",

keywords = "Classification, Compression, Learning",

author = "Chamain, {Lahiru D.} and Cheung, {Sen Ching Samson} and Zhi Ding",

note = "Publisher Copyright: {\textcopyright} 2019 IEEE.; 2019 IEEE International Conference on Multimedia and Expo, ICME 2019 ; Conference date: 08-07-2019 Through 12-07-2019",

year = "2019",

month = jul,

doi = "10.1109/ICME.2019.00066",

language = "English",

series = "Proceedings - IEEE International Conference on Multimedia and Expo",

pages = "338--343",

booktitle = "Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019",

}

Chamain, LD, Cheung, SCS & Ding, Z 2019, Quannet: Joint image compression and classification over channels with limited bandwidth. in Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019., 8784889, Proceedings - IEEE International Conference on Multimedia and Expo, vol. 2019-July, pp. 338-343, 2019 IEEE International Conference on Multimedia and Expo, ICME 2019, Shanghai, China, 7/8/19. https://doi.org/10.1109/ICME.2019.00066

Quannet: Joint image compression and classification over channels with limited bandwidth. / Chamain, Lahiru D.; Cheung, Sen Ching Samson; Ding, Zhi.
Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019. 2019. p. 338-343 8784889 (Proceedings - IEEE International Conference on Multimedia and Expo; Vol. 2019-July).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Quannet

T2 - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019

AU - Chamain, Lahiru D.

AU - Cheung, Sen Ching Samson

AU - Ding, Zhi

PY - 2019/7

Y1 - 2019/7

N2 - The performance of cloud based image classification depends critically on its allocated bandwidth. Traditional data compression methods can negatively impact classification accuracy under limited bandwidth. We investigate the design of bandwidth efficient quantization for image encoding and compression with minimum classification accuracy loss. This work develops a simple neural network framework for joint quantization and classification. The proposed 'QuanNet' can optimize the quantization intervals of JPEG2000 encoder to minimize the classification loss. We show that our quantizer optimization can achieve significant accuracy improvement for a given channel bandwidth. Similarly, significant bandwidth can be saved to achieve a desired accuracy for cloud based image classification.

AB - The performance of cloud based image classification depends critically on its allocated bandwidth. Traditional data compression methods can negatively impact classification accuracy under limited bandwidth. We investigate the design of bandwidth efficient quantization for image encoding and compression with minimum classification accuracy loss. This work develops a simple neural network framework for joint quantization and classification. The proposed 'QuanNet' can optimize the quantization intervals of JPEG2000 encoder to minimize the classification loss. We show that our quantizer optimization can achieve significant accuracy improvement for a given channel bandwidth. Similarly, significant bandwidth can be saved to achieve a desired accuracy for cloud based image classification.

KW - Classification

KW - Compression

KW - Learning

UR - http://www.scopus.com/inward/record.url?scp=85070944487&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85070944487&partnerID=8YFLogxK

U2 - 10.1109/ICME.2019.00066

DO - 10.1109/ICME.2019.00066

M3 - Conference contribution

AN - SCOPUS:85070944487

T3 - Proceedings - IEEE International Conference on Multimedia and Expo

SP - 338

EP - 343

BT - Proceedings - 2019 IEEE International Conference on Multimedia and Expo, ICME 2019

Y2 - 8 July 2019 through 12 July 2019

ER -

Quannet: Joint image compression and classification over channels with limited bandwidth

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this