Reduced-complexity binary-weight-coded associative memories

Hooman Jarollahi; Naoya Onizawa; Vincent Gripon; Warren J. Gross

doi:10.1109/ICASSP.2013.6638110

Reduced-complexity binary-weight-coded associative memories

Hooman Jarollahi, Naoya Onizawa, Vincent Gripon, Warren J. Gross

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

16 Citations (Scopus)

Abstract

Associative memories retrieve stored information given partial or erroneous input patterns. Recently, a new family of associative memories based on Clustered-Neural-Networks (CNNs) was introduced that can store many more messages than classical Hopfield-Neural Networks (HNNs). In this paper, we propose hardware architectures of such memories for partial or erroneous inputs. The proposed architectures eliminate winner-take-all modules and thus reduce the hardware complexity by consuming 65% fewer FPGA lookup tables and increase the operating frequency by approximately 1.9 times compared to that of previous work.

Original language	English
Title of host publication	2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings
Pages	2523-2527
Number of pages	5
DOIs	https://doi.org/10.1109/ICASSP.2013.6638110
Publication status	Published - 2013 Oct 18
Externally published	Yes
Event	2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Vancouver, BC, Canada Duration: 2013 May 26 → 2013 May 31

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
ISSN (Print)	1520-6149

Other

Other	2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013
Country/Territory	Canada
City	Vancouver, BC
Period	13/5/26 → 13/5/31

Keywords

Associative Memories
Clustered Neural Networks
Hardware Implementation
Hopfield

ASJC Scopus subject areas

Software
Signal Processing
Electrical and Electronic Engineering

Access to Document

10.1109/ICASSP.2013.6638110

Cite this

Jarollahi, H., Onizawa, N., Gripon, V., & Gross, W. J. (2013). Reduced-complexity binary-weight-coded associative memories. In 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings (pp. 2523-2527). [6638110] (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). https://doi.org/10.1109/ICASSP.2013.6638110

Reduced-complexity binary-weight-coded associative memories. / Jarollahi, Hooman; Onizawa, Naoya; Gripon, Vincent et al.
2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings. 2013. p. 2523-2527 6638110 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

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

Jarollahi, H, Onizawa, N, Gripon, V & Gross, WJ 2013, Reduced-complexity binary-weight-coded associative memories. in 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings., 6638110, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, pp. 2523-2527, 2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013, Vancouver, BC, Canada, 13/5/26. https://doi.org/10.1109/ICASSP.2013.6638110

Jarollahi H, Onizawa N, Gripon V, Gross WJ. Reduced-complexity binary-weight-coded associative memories. In 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings. 2013. p. 2523-2527. 6638110. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). doi: 10.1109/ICASSP.2013.6638110

@inproceedings{bfddc0ca1b564661a23d38ff94c7607e,

title = "Reduced-complexity binary-weight-coded associative memories",

abstract = "Associative memories retrieve stored information given partial or erroneous input patterns. Recently, a new family of associative memories based on Clustered-Neural-Networks (CNNs) was introduced that can store many more messages than classical Hopfield-Neural Networks (HNNs). In this paper, we propose hardware architectures of such memories for partial or erroneous inputs. The proposed architectures eliminate winner-take-all modules and thus reduce the hardware complexity by consuming 65% fewer FPGA lookup tables and increase the operating frequency by approximately 1.9 times compared to that of previous work.",

keywords = "Associative Memories, Clustered Neural Networks, Hardware Implementation, Hopfield",

author = "Hooman Jarollahi and Naoya Onizawa and Vincent Gripon and Gross, {Warren J.}",

year = "2013",

month = oct,

day = "18",

doi = "10.1109/ICASSP.2013.6638110",

language = "English",

isbn = "9781479903566",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

pages = "2523--2527",

booktitle = "2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings",

note = "2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 ; Conference date: 26-05-2013 Through 31-05-2013",

}

TY - GEN

T1 - Reduced-complexity binary-weight-coded associative memories

AU - Jarollahi, Hooman

AU - Onizawa, Naoya

AU - Gripon, Vincent

AU - Gross, Warren J.

PY - 2013/10/18

Y1 - 2013/10/18

N2 - Associative memories retrieve stored information given partial or erroneous input patterns. Recently, a new family of associative memories based on Clustered-Neural-Networks (CNNs) was introduced that can store many more messages than classical Hopfield-Neural Networks (HNNs). In this paper, we propose hardware architectures of such memories for partial or erroneous inputs. The proposed architectures eliminate winner-take-all modules and thus reduce the hardware complexity by consuming 65% fewer FPGA lookup tables and increase the operating frequency by approximately 1.9 times compared to that of previous work.

AB - Associative memories retrieve stored information given partial or erroneous input patterns. Recently, a new family of associative memories based on Clustered-Neural-Networks (CNNs) was introduced that can store many more messages than classical Hopfield-Neural Networks (HNNs). In this paper, we propose hardware architectures of such memories for partial or erroneous inputs. The proposed architectures eliminate winner-take-all modules and thus reduce the hardware complexity by consuming 65% fewer FPGA lookup tables and increase the operating frequency by approximately 1.9 times compared to that of previous work.

KW - Associative Memories

KW - Clustered Neural Networks

KW - Hardware Implementation

KW - Hopfield

UR - http://www.scopus.com/inward/record.url?scp=84890449236&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890449236&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2013.6638110

DO - 10.1109/ICASSP.2013.6638110

M3 - Conference contribution

AN - SCOPUS:84890449236

SN - 9781479903566

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 2523

EP - 2527

BT - 2013 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013 - Proceedings

T2 - 2013 38th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2013

Y2 - 26 May 2013 through 31 May 2013

ER -

Reduced-complexity binary-weight-coded associative memories

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this