Doc. Ing. Zdeněk Vašíček, Ph.D.

MRÁZEK Vojtěch, SARWAR Syed Shakib, SEKANINA Lukáš, VAŠÍČEK Zdeněk and ROY Kaushik. Design of Power-Efficient Approximate Multipliers for Approximate Artificial Neural Networks. In: Proceedings of the IEEE/ACM International Conference on Computer-Aided Design. Austin, TX: Association for Computing Machinery, 2016, pp. 811-817. ISBN 978-1-4503-4466-1.
Publication language:english
Original title:Design of Power-Efficient Approximate Multipliers for Approximate Artificial Neural Networks
Title (cs):Návrh nízkopříkonových aproximačních násobiček pro aproximační neuronové sítě
Pages:811-817
Proceedings:Proceedings of the IEEE/ACM International Conference on Computer-Aided Design
Conference:2016 IEEE / ACM International Conference On Computer Aided Design
Place:Austin, TX, US
Year:2016
ISBN:978-1-4503-4466-1
Publisher:Association for Computing Machinery
Files: 
+Type Name Title +Size Last modified
iconiccad_nn.pdf741 KB2016-10-05 16:24:00
^ Select all
With selected:
Keywords
Approximate computing, Neural networks, Logic synthesis, Low power, Genetic programming
Annotation
Artificial neural networks (NN) have shown a significant promise in difficult tasks like image classification or speech recognition. Even well-optimized hardware implementations of digital NNs show significant power consumption. It is mainly due to non-uniform pipeline structures and inherent redundancy of numerous arithmetic operations that have to be performed to produce each single output vector.  This paper provides a methodology for the design of well-optimized power-efficient NNs with a uniform structure suitable for hardware implementation. An error resilience analysis was performed in order to determine key constraints for the design of approximate multipliers that are employed in the resulting structure of NN. By means of a search based approximation method, approximate multipliers showing desired tradeoffs between the accuracy and implementation cost were created. Resulting approximate NNs, containing the approximate multipliers, were evaluated using standard benchmarks (MNIST dataset) and a real-world classification problem of Street-View House Numbers. Significant improvement in power efficiency was obtained in both cases with respect to regular NNs. In some cases, 91% power reduction of multiplication led to classification accuracy degradation of less than 2.80%. Moreover, the paper showed the capability of the back propagation learning algorithm to adapt with NNs containing the approximate multipliers. 
BibTeX:
@INPROCEEDINGS{
   author = {Vojt{\v{e}}ch Mr{\'{a}}zek and Shakib Syed Sarwar and
	Luk{\'{a}}{\v{s}} Sekanina and Zden{\v{e}}k
	Va{\v{s}}{\'{i}}{\v{c}}ek and Kaushik Roy},
   title = {Design of Power-Efficient Approximate Multipliers for
	Approximate Artificial Neural Networks},
   pages = {811--817},
   booktitle = {Proceedings of the IEEE/ACM International Conference on
	Computer-Aided Design},
   year = {2016},
   location = {Austin, TX, US},
   publisher = {Association for Computing Machinery},
   ISBN = {978-1-4503-4466-1},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php.en.iso-8859-2?id=11142}
}

Your IPv4 address: 54.224.197.86
Switch to IPv6 connection

DNSSEC [dnssec]