Split and shift methodology on cellular processor arrays: area saving versus time penalty

This paper addresses the so-called split and shift methodology. This methodology deals with the implementation of kernels of sizes that go above the physically implemented resources local connections and weighting circuits on synchronous cellular processor arrays CPA, including the realization of large neighborhood operations and/or the reduction of the available hardware in order to drop the area consumption. Two main goals are pursued in the development of the methodology, namely: 1 minimum penalty at processing time and 2 absolutely no penalty at functional level. The paper presents different techniques and guidelines for the methodology application and introduces a Figure of Merit to evaluate them by relating area gains with time penalty. This, along with a kernel shape analysis, led us to propose more adequate configurations of weighting circuits and to justify the classical choice of North-East-West-South connectivity. To validate the methodology, we realize several estimates over actual physical implementations, and we propose the realization over CPAs of the spin filters, scale invariant feature transform and speeded-up robust features algorithms. A more in-depth trade-off analysis is realized over the implementation of the pixel level snakes algorithm. Copyright © 2012 John Wiley & Sons, Ltd.

[1]  Victor M. Brea,et al.  A binary‐based on‐chip CNN solution for pixel‐level snakes , 2006, Int. J. Circuit Theory Appl..

[2]  Chung-Yu Wu,et al.  A low power design on diffusive interconnection large-neighborhood cellular nonlinear network for giga-scale system application , 2004, Proceedings of the 2004 11th IEEE International Conference on Electronics, Circuits and Systems, 2004. ICECS 2004..

[3]  Piotr Dudek.  Implementation of SIMD vision chip with 128/spl times/128 array of analogue processing elements , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[4]  D. Cabello,et al.  Template-oriented hardware design based on shape analysis of 2D CNN operators in CNN template libraries and applications , 2008, 2008 11th International Workshop on Cellular Neural Networks and Their Applications.

[5]  Tamás Roska,et al.  An optical CNN implementation with stored programmability , 2000, 2000 IEEE International Symposium on Circuits and Systems. Emerging Technologies for the 21st Century. Proceedings (IEEE Cat No.00CH36353).

[6]  Jos Nijhuis,et al.  Transformational DT-CNN design from morphological specifications , 1998 .

[7]  Kari Halonen,et al.  Dense CMOS implementation of a binary‐programmable cellular neural network , 2006, Int. J. Circuit Theory Appl..

[8]  Klaus D. McDonald-Maier,et al.  Exploring Integral Image Word Length Reduction Techniques for SURF Detector , 2009, 2009 Second International Conference on Computer and Electrical Engineering.

[9]  Jordi Albo-Canals,et al.  Robot guiding with obstacle avoidance algorithm for uncertain enviroments based on DTCNN , 2010, The 2010 International Joint Conference on Neural Networks (IJCNN).

[10]  Victor M. Brea,et al.  A one-quadrant discrete-time cellular neural network architecture for pixel-level snakes: B/W processing , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[11]  Mika Laiho,et al.  Cellular nanoscale network cell with memristors for local implication logic and synapses , 2010, Proceedings of 2010 IEEE International Symposium on Circuits and Systems.

[12]  Piotr Dudek,et al.  A general-purpose processor-per-pixel analog SIMD vision chip , 2005, IEEE Transactions on Circuits and Systems I: Regular Papers.

[13]  Chung-Yu Wu,et al.  A new compact neuron-bipolar junction transistor (/spl nu/BJT) cellular neural network (CNN) structure with programmable large neighborhood symmetric templates for image processing , 2001 .

[14]  F. Zou,et al.  A learning algorithm for time-discrete cellular neural networks , 1991, [Proceedings] 1991 IEEE International Joint Conference on Neural Networks.

[15]  D. Cabello,et al.  Verification of Split&Shift techniques for CNN hardware reduction , 2007, 2007 18th European Conference on Circuit Theory and Design.

[16]  Ángel Rodríguez-Vázquez,et al.  ACE16k: the third generation of mixed-signal SIMD-CNN ACE chips toward VSoCs , 2004, IEEE Transactions on Circuits and Systems I: Regular Papers.

[17]  Chung-Yu Wu,et al.  The Design and Analysis of a CMOS Low-Power Large-Neighborhood CNN With Propagating Connections , 2009, IEEE Transactions on Circuits and Systems I: Regular Papers.

[18]  Ángel Rodríguez-Vázquez,et al.  Image filtering by reduced kernels exploiting kernel structure and focal-plane averaging , 2011, 2011 20th European Conference on Circuit Theory and Design (ECCTD).

[19]  Victor M. Brea,et al.  CNN Implementation of Spin Filters for Electronic Speckle Pattern Interferometry Applications , 2007, 2007 IEEE International Symposium on Circuits and Systems.

[20]  David G. Lowe,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004, International Journal of Computer Vision.

[21]  Chung-Yu Wu,et al.  The quantum-dot large-neighborhood cellular nonlinear network (QLN-CNN) in nanotechnology , 2001, Proceedings of the 2001 1st IEEE Conference on Nanotechnology. IEEE-NANO 2001 (Cat. No.01EX516).

[22]  Lambert Spaanenburg,et al.  Optimal decomposition of large-neighborhood CNN templates , 2004 .

[23]  Tamás Roska,et al.  Various implementations of topographic, sensory, cellular wave computers , 2005, 2005 IEEE International Symposium on Circuits and Systems.

[24]  Ákos Zarándy,et al.  2D operators on topographic and non‐topographic architectures—implementation, efficiency analysis, and architecture selection methodology , 2011, Int. J. Circuit Theory Appl..

[25]  Yifan He,et al.  From Xetal-II to Xetal-Pro: On the Road Toward an Ultralow-Energy and High-Throughput SIMD Processor , 2011, IEEE Transactions on Circuits and Systems for Video Technology.

[26]  Leon O. Chua,et al.  Image processing techniques for cellular neural network hardware , 1997 .

[27]  Leon O. Chua,et al.  The CNN paradigm , 1993 .

[28]  L. Spaanenburg,et al.  Efficient DTCNN implementations for large-neighborhood functions , 1998, 1998 Fifth IEEE International Workshop on Cellular Neural Networks and their Applications. Proceedings (Cat. No.98TH8359).

[29]  Kari Halonen,et al.  3-neighborhood motion estimation in CNN silicon architectures , 2004, 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512).

[30]  L. Spaanenburg,et al.  Vein Feature Extraction Using DT-CNNs , 2006, 2006 10th International Workshop on Cellular Neural Networks and Their Applications.

[31]  J.A.K. Suykens,et al.  Fast an Robust Face Tracking for CNN chips: application to wheelchair driving , 2006, 2006 10th International Workshop on Cellular Neural Networks and Their Applications.

[32]  H. Harrer,et al.  New test results of a 4 by 4 discrete-time cellular neural network chip , 1992, CNNA '92 Proceedings Second International Workshop on Cellular Neural Networks and Their Applications.

[33]  Q Yu,et al.  Spin filtering processes and automatic extraction of fringe centerlines in digitial interferometric patterns. , 1988, Applied optics.

[34]  Csaba Rekeczky,et al.  Implementation of a pixel-level snake algorithm on a CNNUM-based chip set architecture , 2004, IEEE Transactions on Circuits and Systems I: Regular Papers.

[35]  V. Bonaiuto,et al.  Time division digital programmable OTA for cellular neural networks , 2005, Proceedings of the 2005 European Conference on Circuit Theory and Design, 2005..

[36]  Elisa Ricci,et al.  Cellular Neural Networks With Virtual Template Expansion for Retinal Vessel Segmentation , 2007, IEEE Transactions on Circuits and Systems II: Express Briefs.

[37]  Victor M. Brea,et al.  An efficient FPGA implementation of a DT-CNN for small image gray-scale pre-processing , 2009, 2009 European Conference on Circuit Theory and Design.

[38]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[39]  Kari Halonen,et al.  An analog array processor hardware realization with multiple new features , 2002, Proceedings of the 2002 International Joint Conference on Neural Networks. IJCNN'02 (Cat. No.02CH37290).

[40]  S. Tokes,et al.  The first version of the optical programmable array/analogic computer (POAC) template library , 2003, 2003 46th Midwest Symposium on Circuits and Systems.

[41]  A. Blug,et al.  Visual inspection of metal objects by using Cellular Neural Networks , 2006, 2006 10th International Workshop on Cellular Neural Networks and Their Applications.

[42]  Kari Halonen,et al.  High density VLSI implementation of a bipolar CNN with reduced programmability , 2004, 2004 IEEE International Symposium on Circuits and Systems (IEEE Cat. No.04CH37512).

[43]  A. Zarandy,et al.  Morphological operators on the CNN Universal Machine , 1996, 1996 Fourth IEEE International Workshop on Cellular Neural Networks and their Applications Proceedings (CNNA-96).

[44]  Piotr Dudek,et al.  Implementation of an asynchronous cellular logic network as a co-processor for a general-purpose massively parallel array , 2007, 2007 18th European Conference on Circuit Theory and Design.

[45]  Chung-Yu Wu,et al.  A new compact programmable /spl nu/BJT cellular neural network structure with adjustable neighborhood layers for image processing , 1999, ICECS'99. Proceedings of ICECS '99. 6th IEEE International Conference on Electronics, Circuits and Systems (Cat. No.99EX357).

[46]  Victor M. Brea,et al.  A binary-based on-chip CNN solution for pixel-level snakes: Research Articles , 2006 .

[47]  Luc Van Gool,et al.  Speeded-Up Robust Features (SURF) , 2008, Comput. Vis. Image Underst..

[48]  L. Spaanenburg,et al.  License plate recognition using DTCNNs , 1998, 1998 Fifth IEEE International Workshop on Cellular Neural Networks and their Applications. Proceedings (Cat. No.98TH8359).

[49]  Jorge Fernandez-Berni,et al.  Accurate design of a MOS-based resistive network for time-controlled diffusion filtering , 2009, 2009 European Conference on Circuit Theory and Design.

[50]  Xose Manuel Pardo,et al.  Cellular neural networks and active contours: a tool for image segmentation , 2003, Image Vis. Comput..