TY - JOUR
T1 - Hardware acceleration for neuromorphic vision algorithms
AU - Al Maashri, Ahmed
AU - Cotter, Matthew
AU - Chandramoorthy, Nandhini
AU - DeBole, Michael
AU - Yu, Chi Li
AU - Narayanan, Vijaykrishnan
AU - Chakrabarti, Chaitali
N1 - Funding Information:
Acknowledgments The authors would like to thank the reviewers for their valuable comments and suggestions. The authors would like to thank Yang Xiao, Penn State, for his role in developing the inter-FPGA communication for the prototyping platform. Also, the authors would like to thank Jim Mutch, MIT for his help in providing the most up to date implementation of the HMAX model. This work was funded in part by DARPA’s NeoVision 2 program, and NSF Awards 1147388, 0916887, 0903432. Ahmed Al Maashri is sponsored by a scholarship from the Government of Oman.
PY - 2013/2
Y1 - 2013/2
N2 - Neuromorphic vision algorithms are biologically inspired models that follow the processing that takes place in the primate visual cortex. Despite their efficiency and robustness, the complexity of these algorithms results in reduced performance when executed on general purpose processors. This paper proposes an application-specific system for accelerating a neuromorphic vision system for object recognition. The system is based on HMAX, a biologically-inspired model of the visual cortex. The neuromorphic accelerators are validated on a multi-FPGA system. Results show that the neuromorphic accelerators are 13.8× (2.6×) more power efficient when compared to CPU (GPU) implementation.
AB - Neuromorphic vision algorithms are biologically inspired models that follow the processing that takes place in the primate visual cortex. Despite their efficiency and robustness, the complexity of these algorithms results in reduced performance when executed on general purpose processors. This paper proposes an application-specific system for accelerating a neuromorphic vision system for object recognition. The system is based on HMAX, a biologically-inspired model of the visual cortex. The neuromorphic accelerators are validated on a multi-FPGA system. Results show that the neuromorphic accelerators are 13.8× (2.6×) more power efficient when compared to CPU (GPU) implementation.
KW - Domain-specific acceleration Power efficiency Neuromorphic systems
UR - http://www.scopus.com/inward/record.url?scp=84892810283&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84892810283&partnerID=8YFLogxK
U2 - 10.1007/s11265-012-0699-x
DO - 10.1007/s11265-012-0699-x
M3 - Article
AN - SCOPUS:84892810283
SN - 1939-8018
VL - 70
SP - 163
EP - 175
JO - Journal of Signal Processing Systems
JF - Journal of Signal Processing Systems
IS - 2
ER -