System-on-chip for biologically inspired vision applications

Sungho Park, Ahmed Al Maashri, Kevin M. Irick, Aarti Chandrashekhar, Matthew Cotter, Nandhini Chandramoorthy, Michael Debole, Vijaykrishnan Narayanan

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Neuromorphic vision algorithms are biologically-inspired computational models of the primate visual pathway. They promise robustness, high accuracy, and high energy efficiency in advanced image processing applications. Despite these potential benefits, the realization of neuromorphic algorithms typically exhibit low performance even when executed on multi-core CPU and GPU platforms. This is due to the disparity in the computational modalities prominent in these algorithms and those modalities most exploited in contemporary computer architectures. In essence, acceleration of neuromorphic algorithms requires adherence to specific computational and communicational requirements. This paper discusses these requirements and proposes a framework for mapping neuromorphic vision applications on a System-on-Chip, SoC. A neuromorphic object detection and recognition on a multi-FPGA platform is presented with performance and power efficiency comparisons to CMP and GPU implementations.

Original languageEnglish
Pages (from-to)71-95
Number of pages25
JournalIPSJ Transactions on System LSI Design Methodology
Volume5
DOIs
Publication statusPublished - 2012

Fingerprint

Computer architecture
Object recognition
Program processors
Energy efficiency
Field programmable gate arrays (FPGA)
Image processing
System-on-chip
Graphics processing unit
Object detection
Primates

Keywords

  • Dataflow process networks
  • Neuromorphic vision
  • Object recognition
  • System-on-chip
  • Visual saliency

ASJC Scopus subject areas

  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Park, S., Al Maashri, A., Irick, K. M., Chandrashekhar, A., Cotter, M., Chandramoorthy, N., ... Narayanan, V. (2012). System-on-chip for biologically inspired vision applications. IPSJ Transactions on System LSI Design Methodology, 5, 71-95. https://doi.org/10.2197/ipsjtsldm.5.71

System-on-chip for biologically inspired vision applications. / Park, Sungho; Al Maashri, Ahmed; Irick, Kevin M.; Chandrashekhar, Aarti; Cotter, Matthew; Chandramoorthy, Nandhini; Debole, Michael; Narayanan, Vijaykrishnan.

In: IPSJ Transactions on System LSI Design Methodology, Vol. 5, 2012, p. 71-95.

Research output: Contribution to journalArticle

Park, S, Al Maashri, A, Irick, KM, Chandrashekhar, A, Cotter, M, Chandramoorthy, N, Debole, M & Narayanan, V 2012, 'System-on-chip for biologically inspired vision applications', IPSJ Transactions on System LSI Design Methodology, vol. 5, pp. 71-95. https://doi.org/10.2197/ipsjtsldm.5.71
Park, Sungho ; Al Maashri, Ahmed ; Irick, Kevin M. ; Chandrashekhar, Aarti ; Cotter, Matthew ; Chandramoorthy, Nandhini ; Debole, Michael ; Narayanan, Vijaykrishnan. / System-on-chip for biologically inspired vision applications. In: IPSJ Transactions on System LSI Design Methodology. 2012 ; Vol. 5. pp. 71-95.
@article{f7c5c6126fb7439fb8c6d3d3fd586d23,
title = "System-on-chip for biologically inspired vision applications",
abstract = "Neuromorphic vision algorithms are biologically-inspired computational models of the primate visual pathway. They promise robustness, high accuracy, and high energy efficiency in advanced image processing applications. Despite these potential benefits, the realization of neuromorphic algorithms typically exhibit low performance even when executed on multi-core CPU and GPU platforms. This is due to the disparity in the computational modalities prominent in these algorithms and those modalities most exploited in contemporary computer architectures. In essence, acceleration of neuromorphic algorithms requires adherence to specific computational and communicational requirements. This paper discusses these requirements and proposes a framework for mapping neuromorphic vision applications on a System-on-Chip, SoC. A neuromorphic object detection and recognition on a multi-FPGA platform is presented with performance and power efficiency comparisons to CMP and GPU implementations.",
keywords = "Dataflow process networks, Neuromorphic vision, Object recognition, System-on-chip, Visual saliency",
author = "Sungho Park and {Al Maashri}, Ahmed and Irick, {Kevin M.} and Aarti Chandrashekhar and Matthew Cotter and Nandhini Chandramoorthy and Michael Debole and Vijaykrishnan Narayanan",
year = "2012",
doi = "10.2197/ipsjtsldm.5.71",
language = "English",
volume = "5",
pages = "71--95",
journal = "IPSJ Transactions on System LSI Design Methodology",
issn = "1882-6687",
publisher = "Information Processing Society of Japan",

}

TY - JOUR

T1 - System-on-chip for biologically inspired vision applications

AU - Park, Sungho

AU - Al Maashri, Ahmed

AU - Irick, Kevin M.

AU - Chandrashekhar, Aarti

AU - Cotter, Matthew

AU - Chandramoorthy, Nandhini

AU - Debole, Michael

AU - Narayanan, Vijaykrishnan

PY - 2012

Y1 - 2012

N2 - Neuromorphic vision algorithms are biologically-inspired computational models of the primate visual pathway. They promise robustness, high accuracy, and high energy efficiency in advanced image processing applications. Despite these potential benefits, the realization of neuromorphic algorithms typically exhibit low performance even when executed on multi-core CPU and GPU platforms. This is due to the disparity in the computational modalities prominent in these algorithms and those modalities most exploited in contemporary computer architectures. In essence, acceleration of neuromorphic algorithms requires adherence to specific computational and communicational requirements. This paper discusses these requirements and proposes a framework for mapping neuromorphic vision applications on a System-on-Chip, SoC. A neuromorphic object detection and recognition on a multi-FPGA platform is presented with performance and power efficiency comparisons to CMP and GPU implementations.

AB - Neuromorphic vision algorithms are biologically-inspired computational models of the primate visual pathway. They promise robustness, high accuracy, and high energy efficiency in advanced image processing applications. Despite these potential benefits, the realization of neuromorphic algorithms typically exhibit low performance even when executed on multi-core CPU and GPU platforms. This is due to the disparity in the computational modalities prominent in these algorithms and those modalities most exploited in contemporary computer architectures. In essence, acceleration of neuromorphic algorithms requires adherence to specific computational and communicational requirements. This paper discusses these requirements and proposes a framework for mapping neuromorphic vision applications on a System-on-Chip, SoC. A neuromorphic object detection and recognition on a multi-FPGA platform is presented with performance and power efficiency comparisons to CMP and GPU implementations.

KW - Dataflow process networks

KW - Neuromorphic vision

KW - Object recognition

KW - System-on-chip

KW - Visual saliency

UR - http://www.scopus.com/inward/record.url?scp=84864918550&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84864918550&partnerID=8YFLogxK

U2 - 10.2197/ipsjtsldm.5.71

DO - 10.2197/ipsjtsldm.5.71

M3 - Article

AN - SCOPUS:84864918550

VL - 5

SP - 71

EP - 95

JO - IPSJ Transactions on System LSI Design Methodology

JF - IPSJ Transactions on System LSI Design Methodology

SN - 1882-6687

ER -