{"title":"A topological description of the state space of a cellular neural network","authors":"P. Civalleri, M. Gilli","doi":"10.1109/CNNA.1994.381697","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381697","url":null,"abstract":"The structure of the state space of cellular neural networks is investigated by putting the invariant manifolds of the fixed points of networks having the maximum number of equilibria in one-to-one correspondence with the cells of various orders of an n-cube (where n is the dimension of the state space) and of its dual. It is shown that the set of such networks is non-void for any template structure and that bifurcations of equilibria correspond to either vanishing or shrinking of cells in both complexes. Both topological representations provide an intuitive description of the geometrical features underlying the network dynamics.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121446796","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"An approach to local diffusion and global propagation in 1-dim. cellular neural networks","authors":"Patrick Thiran, G. Setti","doi":"10.1109/CNNA.1994.381653","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381653","url":null,"abstract":"Summary form only given, as follows. We study the phenomena of local diffusion and global propagation in a one-dimensional CNN described by the space-invariant A-template A = [A/sub -1/ A/sub 0/ A/sub 1/]. Roughly speaking, a CNN behaves in a local diffusion mode when two distant cells do not influence each other if the states of a number r of adjacent cells located between these two cells have reached some value. It behaves in a global propagation mode otherwise, i.e. when one of these two cells can always influence the other one, whatever the value of the state of r adjacent cells located in between these two cells. We can then compute the values of the template parameters for which the CNN has one of these behaviors. The distinction between these two methods of information processing is a radical one that has many practical consequences on: stability; the influence of boundary conditions; the dependence of the number of stable equilibria on the number of cells; the existence of limit cycles; and on the lengths of transients. For example, we can prove that the number of stable equilibria grows exponentially with the number of cells if and only if the CNN has a local diffusion behavior. If it operates in a global propagation mode, this is no longer true, but periodic solutions (one of which can be explicitly computed) are then present for some types of boundary conditions.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128643701","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A massively parallel approach to cellular neural networks image processing","authors":"Giovanni Adorni, V. D'Andrea, G. Destri","doi":"10.1109/CNNA.1994.381638","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381638","url":null,"abstract":"Low-level image processing is a critical phase, since the results of following \"more intelligent\" steps depend on the output quality of the first processing stages. In this work we present an edge detection filtering algorithm, strictly oriented to enhance the edge of some objects, in a typical real image noisy context, using some \"a priori\" known characteristics. We describe also an application of this algorithm to the analysis of road images, where the goal is the enhancement of traffic signs.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114576260","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Automatic design of cellular neural networks by means of genetic algorithms: finding a feature detector","authors":"F. Dellaert, J. Vandewalle","doi":"10.1109/CNNA.1994.381681","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381681","url":null,"abstract":"The paper aims to examine the use of genetic algorithms to optimize subsystems of cellular neural network architectures. The application at hand is character recognition: the aim is to evolve an optimal feature detector in order to aid a conventional classifier network to generalize across different fonts. To this end, a performance function and a genetic encoding for a feature detector are presented. An experiment is described where an optimal feature detector is indeed found by the genetic algorithm.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127831573","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Distributed neural control for Markov decision processes in hierarchic communication networks","authors":"A. Murgu","doi":"10.1109/CNNA.1994.381664","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381664","url":null,"abstract":"We consider the problem of traffic control in a communication network hierarchically organized in order to handle different flow streams. Such a problem exhibits a partitioned control structure and is usually solved using preplanned routing mechanisms for slow varying flows. Here, we consider a Markov decision process modelling of the entities requiring admission and routing through the network under a setting allowing the slotted processing, extending a previous research (Murgu et al., 1994) done for a queueing system modelled as Brownian control problem. Using an idea of certainty equivalence principle, we map the local control scheme into a recurrent adaptive algorithm in which a learning mechanism is included. Finally, a numerical experiment with medium sized network is considered and the main conclusions are reported.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126636749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
W. Jansen, R. van Drunen, L. Spaanenburg, J. Nijhuis
{"title":"The AD/sup 2/ microcontroller extension for artificial neural networks","authors":"W. Jansen, R. van Drunen, L. Spaanenburg, J. Nijhuis","doi":"10.1109/CNNA.1994.381651","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381651","url":null,"abstract":"The lack of dynamic precision has so far hindered the introduction of real-time neural data processing in the popular 8-bit microcontroller market. It is discussed how, within the framework of micro embedding and soft programming, neural capability can be introduced while allowing for the original application software.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124944503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"High performance digitally programmable CNN chip with discrete templates","authors":"F. Sargeni, V. Bonaiuto","doi":"10.1109/CNNA.1994.381705","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381705","url":null,"abstract":"In this paper a high performance VLSI implementation of a 3/spl times/3 Digitally Programmable Cellular Neural Network with discrete templates is presented. This chip, manufactured and successfully tested, gives an efficient solution to the hardware implementation of the Cellular Neural Networks. Moreover this chip can be connected to others to carry out very large CNN arrays. This implementation covers the 66% of the available one-neighborhood fixed templates for image processing applications.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124369250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Hyperchaos, clustering and cooperative phenomena in CNN arrays composed of chaotic circuits","authors":"M. Ogorzałek, A. Dabrowski, W. Dąbrowski","doi":"10.1109/CNNA.1994.381660","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381660","url":null,"abstract":"This paper presents a study of complex dynamic phenomena in arrays composed of interacting chaotic circuits. Such arrays are thought of as a new paradigm for signal processing. Depending on the connection strength between the cells the array can show disorganised hyperchaotic behavior or organised in a specific way coherent behavior which might be useful in information processing. Patterns of behavior depending on coupling values are studied in this paper. Chua's circuits are taken as standard chaotic cells.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123527840","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Random variations in CNN templates: theoretical models and empirical studies","authors":"Bertram E. Shi, S. Wendsche, T. Roska, L. Chua","doi":"10.1109/CNNA.1994.381684","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381684","url":null,"abstract":"This paper studies the performance of binary image processing CNN templates when the actual template values at each cell are allowed to vary from their nominal values. We examine the validity of one plausible measure of the robustness to random template variations: the minimum absolute value of the current into the capacitor taken over all possible binary state patterns divided by the norm of the template elements. While this measure can be proven to be a valid indicator of robustness for linear threshold templates, its predictive power on the more dynamically complex CCD template is mixed. In some cases, an estimate of the error rate based upon this measure matches remarkably well with the results of numerical simulations. In others, this measure of robustness predicts that one template is more robust than another, while numerical simulations indicate that the opposite is true.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"772 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124602958","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A reconfigurable architecture mapping multilayer CNN paradigms","authors":"L. Raffo, S. Sabatini, G. Bisio","doi":"10.1109/CNNA.1994.381643","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381643","url":null,"abstract":"A digital VLSI implementation of linear template cellular neural nets (CNNs) is presented. A reconfigurable architecture is organized as 12 layers of 64/spl times/64 cells. The CNNs are reformulated introducing sets of generalized cloning templates to enucleate more sharply the structure of both intra- and inter-layer cooperative computations. In this way it is possible to develop CNN algorithms for complex vision machine tasks. Various applications are considered in edge and connected component detection and in texture segregation.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129524646","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}