DocumentCode
1749052
Title
Comparative study of implementing ICNNs on FPGAs
Author
Lim, Ai Boon ; Rajapakse, Jagath C. ; Omondi, Amos R.
Author_Institution
Sch. of Comput. Eng., Nanyang Technol. Univ., Singapore
Volume
1
fYear
2001
fDate
2001
Firstpage
177
Abstract
Hardware implementation of two independent component neural networks, one that maximizes the mutual-information between input and output and one that minimizes the divergence at the output, is investigated. Table lookup and computer arithmetic implementations with field-programmable gate arrays are compared. The lookup table approach is superior, in terms of both cost and performance. The network that minimizes output-divergence is less costly to implement than that which maximizes mutual information
Keywords
field programmable gate arrays; neural net architecture; performance evaluation; principal component analysis; table lookup; field-programmable gate arrays; hardware; independent component analysis; independent component neural networks; output-divergence; statistical independence; table lookup; Application software; Blind source separation; Computer architecture; Costs; Field programmable gate arrays; Independent component analysis; Mutual information; Neural network hardware; Neural networks; Neurons;
fLanguage
English
Publisher
ieee
Conference_Titel
Neural Networks, 2001. Proceedings. IJCNN '01. International Joint Conference on
Conference_Location
Washington, DC
ISSN
1098-7576
Print_ISBN
0-7803-7044-9
Type
conf
DOI
10.1109/IJCNN.2001.939013
Filename
939013
Link To Document