CNNOPT: Learning dynamics and CNN chip-specific robustness
- 1 August 2006
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 21650144,p. 1-6
Abstract
A method is presented that unifies previous approaches with the aim of learning new templates with the ability to tune cellular nonlinear network (CNN) templates to individual chip instances in a global optimization framework. The proposed method is built on earlier approaches extending them in three main aspects. First, hardware parameters of the CNN chip are included in the optimization that opens the way to run templates so far believed to be very unstable on chip. Second, a novel global optimization algorithm is used that improves learning speed significantly. Third, the whole method is presented as a new Matlab toolbox so that the only task of the CNN algorithm designer is to formulate the operation to be learned as a training set of the optimization process. Training set design is the most crucial issue of this approach, thus basic rules for the design of training sets are presented. Examples are given in order to illustrate the design issues. We believe that the proposed method can be a valuable tool to find new CNN templates and robustly implement them on chipKeywords
This publication has 11 references indexed in Scilit:
- Topographic cellular active contour techniques: theory, implementations and comparisonsInternational Journal of Circuit Theory and Applications, 2006
- Learning of spatiotemporal behaviour in cellular neural networksInternational Journal of Circuit Theory and Applications, 2006
- Bi-i: a standalone ultra high speed cellular vision systemIEEE Circuits and Systems Magazine, 2005
- ACE16k: The Third Generation of Mixed-Signal SIMD-CNN ACE Chips Toward VSoCsIEEE Transactions on Circuits and Systems I: Regular Papers, 2004
- Toward CNN Chip-Specific RobustnessIEEE Transactions on Circuits and Systems I: Regular Papers, 2004
- Morphology and autowave metric on CNN applied to bubble-debris classificationIEEE Transactions on Neural Networks, 2000
- On Stability of Cellular Neural NetworksJournal of Signal Processing Systems, 1999
- Computing with Front Propagation: Active Contour And Skeleton Models In Continuous-Time CNNJournal of Signal Processing Systems, 1999
- The art of CNN template designInternational Journal of Circuit Theory and Applications, 1999
- Minimizing the effects of parameter deviations on cellular neural networksInternational Journal of Circuit Theory and Applications, 1999