A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS

Minkyu Kim, Abinash Mohanty, Deepak Kadetotad, Naveen Suda, Luning Wei, Pooja Saseendran, Xiaofei He, Yu Cao, Jae-sun Seo

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper presents an object detection accelerator that features many-scale (17), many-object (up to 50), multi-class (e.g., face, traffic sign), and high accuracy (average precision of 0.79/0.65 for AFW/BTSD datasets). Employing 10 gradient/color channels, integral features are extracted, and the results of 2,000 simple classifiers for rigid boosted templates are adaptively combined to make a strong classification. By jointly optimizing the algorithm and the hardware architecture, the prototype chip implemented in 65nm CMOS demonstrates real-time object detection of 13-35 frames per second with low power consumption of 22-160mW at 0.58-1.0V supply.

Original languageEnglish (US)
Title of host publicationIEEE International Symposium on Circuits and Systems
Subtitle of host publicationFrom Dreams to Innovation, ISCAS 2017 - Conference Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781467368520
DOIs
StatePublished - Sep 25 2017
Event50th IEEE International Symposium on Circuits and Systems, ISCAS 2017 - Baltimore, United States
Duration: May 28 2017May 31 2017

Other

Other50th IEEE International Symposium on Circuits and Systems, ISCAS 2017
CountryUnited States
CityBaltimore
Period5/28/175/31/17

Fingerprint

Particle accelerators
Traffic signs
Electric power utilization
Classifiers
Color
Hardware
Object detection

Keywords

  • classification
  • low-power
  • machine learning
  • object detection
  • real-time
  • special-purpose accelerator

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Cite this

Kim, M., Mohanty, A., Kadetotad, D., Suda, N., Wei, L., Saseendran, P., ... Seo, J. (2017). A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS. In IEEE International Symposium on Circuits and Systems: From Dreams to Innovation, ISCAS 2017 - Conference Proceedings [8050798] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ISCAS.2017.8050798

A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS. / Kim, Minkyu; Mohanty, Abinash; Kadetotad, Deepak; Suda, Naveen; Wei, Luning; Saseendran, Pooja; He, Xiaofei; Cao, Yu; Seo, Jae-sun.

IEEE International Symposium on Circuits and Systems: From Dreams to Innovation, ISCAS 2017 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017. 8050798.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kim, M, Mohanty, A, Kadetotad, D, Suda, N, Wei, L, Saseendran, P, He, X, Cao, Y & Seo, J 2017, A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS. in IEEE International Symposium on Circuits and Systems: From Dreams to Innovation, ISCAS 2017 - Conference Proceedings., 8050798, Institute of Electrical and Electronics Engineers Inc., 50th IEEE International Symposium on Circuits and Systems, ISCAS 2017, Baltimore, United States, 5/28/17. https://doi.org/10.1109/ISCAS.2017.8050798
Kim M, Mohanty A, Kadetotad D, Suda N, Wei L, Saseendran P et al. A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS. In IEEE International Symposium on Circuits and Systems: From Dreams to Innovation, ISCAS 2017 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc. 2017. 8050798 https://doi.org/10.1109/ISCAS.2017.8050798
Kim, Minkyu ; Mohanty, Abinash ; Kadetotad, Deepak ; Suda, Naveen ; Wei, Luning ; Saseendran, Pooja ; He, Xiaofei ; Cao, Yu ; Seo, Jae-sun. / A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS. IEEE International Symposium on Circuits and Systems: From Dreams to Innovation, ISCAS 2017 - Conference Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017.
@inproceedings{69e7424cad504e88a4a33ea5621d2777,
title = "A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS",
abstract = "This paper presents an object detection accelerator that features many-scale (17), many-object (up to 50), multi-class (e.g., face, traffic sign), and high accuracy (average precision of 0.79/0.65 for AFW/BTSD datasets). Employing 10 gradient/color channels, integral features are extracted, and the results of 2,000 simple classifiers for rigid boosted templates are adaptively combined to make a strong classification. By jointly optimizing the algorithm and the hardware architecture, the prototype chip implemented in 65nm CMOS demonstrates real-time object detection of 13-35 frames per second with low power consumption of 22-160mW at 0.58-1.0V supply.",
keywords = "classification, low-power, machine learning, object detection, real-time, special-purpose accelerator",
author = "Minkyu Kim and Abinash Mohanty and Deepak Kadetotad and Naveen Suda and Luning Wei and Pooja Saseendran and Xiaofei He and Yu Cao and Jae-sun Seo",
year = "2017",
month = "9",
day = "25",
doi = "10.1109/ISCAS.2017.8050798",
language = "English (US)",
booktitle = "IEEE International Symposium on Circuits and Systems",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - A real-time 17-scale object detection accelerator with adaptive 2000-stage classification in 65nm CMOS

AU - Kim, Minkyu

AU - Mohanty, Abinash

AU - Kadetotad, Deepak

AU - Suda, Naveen

AU - Wei, Luning

AU - Saseendran, Pooja

AU - He, Xiaofei

AU - Cao, Yu

AU - Seo, Jae-sun

PY - 2017/9/25

Y1 - 2017/9/25

N2 - This paper presents an object detection accelerator that features many-scale (17), many-object (up to 50), multi-class (e.g., face, traffic sign), and high accuracy (average precision of 0.79/0.65 for AFW/BTSD datasets). Employing 10 gradient/color channels, integral features are extracted, and the results of 2,000 simple classifiers for rigid boosted templates are adaptively combined to make a strong classification. By jointly optimizing the algorithm and the hardware architecture, the prototype chip implemented in 65nm CMOS demonstrates real-time object detection of 13-35 frames per second with low power consumption of 22-160mW at 0.58-1.0V supply.

AB - This paper presents an object detection accelerator that features many-scale (17), many-object (up to 50), multi-class (e.g., face, traffic sign), and high accuracy (average precision of 0.79/0.65 for AFW/BTSD datasets). Employing 10 gradient/color channels, integral features are extracted, and the results of 2,000 simple classifiers for rigid boosted templates are adaptively combined to make a strong classification. By jointly optimizing the algorithm and the hardware architecture, the prototype chip implemented in 65nm CMOS demonstrates real-time object detection of 13-35 frames per second with low power consumption of 22-160mW at 0.58-1.0V supply.

KW - classification

KW - low-power

KW - machine learning

KW - object detection

KW - real-time

KW - special-purpose accelerator

UR - http://www.scopus.com/inward/record.url?scp=85032692087&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85032692087&partnerID=8YFLogxK

U2 - 10.1109/ISCAS.2017.8050798

DO - 10.1109/ISCAS.2017.8050798

M3 - Conference contribution

AN - SCOPUS:85032692087

BT - IEEE International Symposium on Circuits and Systems

PB - Institute of Electrical and Electronics Engineers Inc.

ER -