A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion

Jintao Zhang; Zhuo Wang; Naveen Verma

doi:10.1109/ISSCC.2015.7063061

A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion

Jintao Zhang, Zhuo Wang, Naveen Verma

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

34 Scopus citations

Abstract

Embedded sensing systems conventionally perform A-to-D conversion followed by signal analysis. In many applications, the analysis of interest is inference (e.g., classification), but the sensor signals involved are too complex to model analytically. Machine learning is gaining prominence because it enables data-driven training of classifiers, overcoming the need for analytical models. This work presents: 1) an algorithmic formulation, where feature extraction and classification are combined into a single matrix, reducing the total multiplications needed, and 2) a matrix-multiplying ADC (MMADC) that enables multiplication of input samples by a programmable matrix. Thus, the MMADC combines feature extraction and classification with data conversion, mitigating the need for further computations. Two systems are demonstrated: an ECG-based cardiac-arrhythmia detector and an image-pixel-based gender detector.

Original language	English (US)
Title of host publication	2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	332-333
Number of pages	2
ISBN (Electronic)	9781479962235
DOIs	https://doi.org/10.1109/ISSCC.2015.7063061
State	Published - Mar 17 2015
Event	2015 62nd IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers - San Francisco, United States Duration: Feb 22 2015 → Feb 26 2015

Publication series

Name	Digest of Technical Papers - IEEE International Solid-State Circuits Conference
Volume	58
ISSN (Print)	0193-6530

Other

Other	2015 62nd IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers
Country/Territory	United States
City	San Francisco
Period	2/22/15 → 2/26/15

All Science Journal Classification (ASJC) codes

Electronic, Optical and Magnetic Materials
Electrical and Electronic Engineering

Access to Document

10.1109/ISSCC.2015.7063061

Cite this

Zhang, J., Wang, Z., & Verma, N. (2015). A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion. In 2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers (pp. 332-333). Article 7063061 (Digest of Technical Papers - IEEE International Solid-State Circuits Conference; Vol. 58). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISSCC.2015.7063061

Zhang, Jintao ; Wang, Zhuo ; Verma, Naveen. / A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion. 2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers. Institute of Electrical and Electronics Engineers Inc., 2015. pp. 332-333 (Digest of Technical Papers - IEEE International Solid-State Circuits Conference).

@inproceedings{d4495c331f7c42a1a28dede0d828a67d,

title = "A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion",

abstract = "Embedded sensing systems conventionally perform A-to-D conversion followed by signal analysis. In many applications, the analysis of interest is inference (e.g., classification), but the sensor signals involved are too complex to model analytically. Machine learning is gaining prominence because it enables data-driven training of classifiers, overcoming the need for analytical models. This work presents: 1) an algorithmic formulation, where feature extraction and classification are combined into a single matrix, reducing the total multiplications needed, and 2) a matrix-multiplying ADC (MMADC) that enables multiplication of input samples by a programmable matrix. Thus, the MMADC combines feature extraction and classification with data conversion, mitigating the need for further computations. Two systems are demonstrated: an ECG-based cardiac-arrhythmia detector and an image-pixel-based gender detector.",

author = "Jintao Zhang and Zhuo Wang and Naveen Verma",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; 2015 62nd IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers ; Conference date: 22-02-2015 Through 26-02-2015",

year = "2015",

month = mar,

day = "17",

doi = "10.1109/ISSCC.2015.7063061",

language = "English (US)",

series = "Digest of Technical Papers - IEEE International Solid-State Circuits Conference",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "332--333",

booktitle = "2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers",

address = "United States",

}

Zhang, J, Wang, Z & Verma, N 2015, A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion. in 2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers., 7063061, Digest of Technical Papers - IEEE International Solid-State Circuits Conference, vol. 58, Institute of Electrical and Electronics Engineers Inc., pp. 332-333, 2015 62nd IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers, San Francisco, United States, 2/22/15. https://doi.org/10.1109/ISSCC.2015.7063061

A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion. / Zhang, Jintao; Wang, Zhuo; Verma, Naveen.
2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers. Institute of Electrical and Electronics Engineers Inc., 2015. p. 332-333 7063061 (Digest of Technical Papers - IEEE International Solid-State Circuits Conference; Vol. 58).

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

TY - GEN

T1 - A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion

AU - Zhang, Jintao

AU - Wang, Zhuo

AU - Verma, Naveen

PY - 2015/3/17

Y1 - 2015/3/17

N2 - Embedded sensing systems conventionally perform A-to-D conversion followed by signal analysis. In many applications, the analysis of interest is inference (e.g., classification), but the sensor signals involved are too complex to model analytically. Machine learning is gaining prominence because it enables data-driven training of classifiers, overcoming the need for analytical models. This work presents: 1) an algorithmic formulation, where feature extraction and classification are combined into a single matrix, reducing the total multiplications needed, and 2) a matrix-multiplying ADC (MMADC) that enables multiplication of input samples by a programmable matrix. Thus, the MMADC combines feature extraction and classification with data conversion, mitigating the need for further computations. Two systems are demonstrated: an ECG-based cardiac-arrhythmia detector and an image-pixel-based gender detector.

AB - Embedded sensing systems conventionally perform A-to-D conversion followed by signal analysis. In many applications, the analysis of interest is inference (e.g., classification), but the sensor signals involved are too complex to model analytically. Machine learning is gaining prominence because it enables data-driven training of classifiers, overcoming the need for analytical models. This work presents: 1) an algorithmic formulation, where feature extraction and classification are combined into a single matrix, reducing the total multiplications needed, and 2) a matrix-multiplying ADC (MMADC) that enables multiplication of input samples by a programmable matrix. Thus, the MMADC combines feature extraction and classification with data conversion, mitigating the need for further computations. Two systems are demonstrated: an ECG-based cardiac-arrhythmia detector and an image-pixel-based gender detector.

UR - http://www.scopus.com/inward/record.url?scp=84940733990&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84940733990&partnerID=8YFLogxK

U2 - 10.1109/ISSCC.2015.7063061

DO - 10.1109/ISSCC.2015.7063061

M3 - Conference contribution

AN - SCOPUS:84940733990

T3 - Digest of Technical Papers - IEEE International Solid-State Circuits Conference

SP - 332

EP - 333

BT - 2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2015 62nd IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers

Y2 - 22 February 2015 through 26 February 2015

ER -

Zhang J, Wang Z, Verma N. A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion. In 2015 IEEE International Solid-State Circuits Conference, ISSCC 2015 - Digest of Technical Papers. Institute of Electrical and Electronics Engineers Inc. 2015. p. 332-333. 7063061. (Digest of Technical Papers - IEEE International Solid-State Circuits Conference). doi: 10.1109/ISSCC.2015.7063061

A matrix-multiplying ADC implementing a machine-learning classifier directly with data conversion

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this