Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381695
Patrick Thiran, M. Hasler
We review some principles for information storage and processing, based on oscillations in dynamical systems. Oscillations and chaos are present in both biological and artificial neurons. A single biological neuron has an oscillatory dynamics, and can generate chaos. At a macroscopic level however, chaos is not created by the dynamics of individual neurons, but by the interaction of large groups of neurons. These macroscopic oscillations are measured by EEG recordings that indicate the presence of chaotic attractors in the brain. Also in the visual cortex, neurons have been found to oscillate in a coherent way depending on the global stimulus. On the other hand, as recurrent artificial neural networks are non linear dynamical systems, it is possible to get different behaviours by adjusting their parameters: convergence toward equilibrium points, toward periodic solutions or chaotic trajectories. In this case, the study of oscillations is more a scientific activity than a goal for storing and processing information. In this paper, however, we explore the possibilities to make use of chaos for information storage.<>
{"title":"Information processing using stable and unstable oscillations: a tutorial","authors":"Patrick Thiran, M. Hasler","doi":"10.1109/CNNA.1994.381695","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381695","url":null,"abstract":"We review some principles for information storage and processing, based on oscillations in dynamical systems. Oscillations and chaos are present in both biological and artificial neurons. A single biological neuron has an oscillatory dynamics, and can generate chaos. At a macroscopic level however, chaos is not created by the dynamics of individual neurons, but by the interaction of large groups of neurons. These macroscopic oscillations are measured by EEG recordings that indicate the presence of chaotic attractors in the brain. Also in the visual cortex, neurons have been found to oscillate in a coherent way depending on the global stimulus. On the other hand, as recurrent artificial neural networks are non linear dynamical systems, it is possible to get different behaviours by adjusting their parameters: convergence toward equilibrium points, toward periodic solutions or chaotic trajectories. In this case, the study of oscillations is more a scientific activity than a goal for storing and processing information. In this paper, however, we explore the possibilities to make use of chaos for information storage.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"172 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":"132482226","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381650
C. Pham, M. Tanaka
Bifurcation and chaotic behaviors which occur in simple looped CMOS circuit with high speed operation are described. The bifurcation and chaotic behavior have been found along with a variation of a sampling clock frequency.<>
{"title":"A novel chaos generator employing CMOS inverter for cellular neural networks","authors":"C. Pham, M. Tanaka","doi":"10.1109/CNNA.1994.381650","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381650","url":null,"abstract":"Bifurcation and chaotic behaviors which occur in simple looped CMOS circuit with high speed operation are described. The bifurcation and chaotic behavior have been found along with a variation of a sampling clock frequency.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"9 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":"121683197","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381673
J.P. Miller, T. Roska, T. Szirányi, K. R. Crounse, L. Chua, L. Nemes
In this paper it is shown how the Cellular Neural Network (CNN) can be used to perform image and volume deblurring, with particular emphases on applications to microscopy. We discuss the basic linear theory of the CNN including issues of stability and template size. It is observed that a CNN with a small template can be used to implement an Infinite Impulse Response filter. It is then shown how general deblurring problems can be addressed with a CNN when the blurring operator is known. The proposed application is to solve the basic 3-D confocal image reconstruction task about the form of the blurring operator, confocal behavior in microscope images can be obtained with only 3-5 acquired image planes. In addition, the stored program capability of the CNN Universal Machine would provide integration of several image processing and detection tasks in the same architecture.<>
{"title":"Deblurring of images by cellular neural networks with applications to microscopy","authors":"J.P. Miller, T. Roska, T. Szirányi, K. R. Crounse, L. Chua, L. Nemes","doi":"10.1109/CNNA.1994.381673","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381673","url":null,"abstract":"In this paper it is shown how the Cellular Neural Network (CNN) can be used to perform image and volume deblurring, with particular emphases on applications to microscopy. We discuss the basic linear theory of the CNN including issues of stability and template size. It is observed that a CNN with a small template can be used to implement an Infinite Impulse Response filter. It is then shown how general deblurring problems can be addressed with a CNN when the blurring operator is known. The proposed application is to solve the basic 3-D confocal image reconstruction task about the form of the blurring operator, confocal behavior in microscope images can be obtained with only 3-5 acquired image planes. In addition, the stored program capability of the CNN Universal Machine would provide integration of several image processing and detection tasks in the same architecture.<<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":"124038961","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381682
A. Schuler, M. Brabec, D. Schubel, J. Nossek
The paper presents an approach to learning, which focuses on finding a set of parameter values taking into account the nonidealities of a specific implementation. Therefore learning is done on a more accurate model of a CMOS cell, and not on the original CNN model proposed by Chua and Yang (1988) and Nossek et al. (1990). This hardware-oriented approach is applied to a current-mode CNN-model based on the full-signal-range model of Rodriguez-Vaazquez et al. (1993) and Espejo (1994), where the dynamic block consists of two current mirrors. It is shown, that a two-quadrant multiplier is sufficient for the multiplication with the template coefficients, by changing the model, further reducing the area consumption. Using a hardware-oriented approach to learning thus not only allows to find template values for a specific VLSI-implementation, but may also lead to further simplifications of CNN-implementations.<>
{"title":"Hardware-oriented learning for cellular neural networks","authors":"A. Schuler, M. Brabec, D. Schubel, J. Nossek","doi":"10.1109/CNNA.1994.381682","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381682","url":null,"abstract":"The paper presents an approach to learning, which focuses on finding a set of parameter values taking into account the nonidealities of a specific implementation. Therefore learning is done on a more accurate model of a CMOS cell, and not on the original CNN model proposed by Chua and Yang (1988) and Nossek et al. (1990). This hardware-oriented approach is applied to a current-mode CNN-model based on the full-signal-range model of Rodriguez-Vaazquez et al. (1993) and Espejo (1994), where the dynamic block consists of two current mirrors. It is shown, that a two-quadrant multiplier is sufficient for the multiplication with the template coefficients, by changing the model, further reducing the area consumption. Using a hardware-oriented approach to learning thus not only allows to find template values for a specific VLSI-implementation, but may also lead to further simplifications of CNN-implementations.<<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":"131164792","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381677
Krzysztof Slot
The paper presents the method of large neighborhood templates realization (i.e. templates with r>1) in a nearest-neighbor connected (i.e. r=1) discrete-time CNN Universal Machine. This is accomplished by decomposing an objective template into a sum of two-dimensional 3/spl times/3 template correlations. An appropriate procedure which ensures a desired circuit operation is given in an algorithmic form.<>
{"title":"Large-neighborhood templates implementation in discrete-time CNN Universal Machine with a nearest-neighbor connection pattern","authors":"Krzysztof Slot","doi":"10.1109/CNNA.1994.381677","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381677","url":null,"abstract":"The paper presents the method of large neighborhood templates realization (i.e. templates with r>1) in a nearest-neighbor connected (i.e. r=1) discrete-time CNN Universal Machine. This is accomplished by decomposing an objective template into a sum of two-dimensional 3/spl times/3 template correlations. An appropriate procedure which ensures a desired circuit operation is given in an algorithmic form.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"64 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":"121588195","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381657
S. Jankowski, R. Wanczuk
The paper presents the nonlinear discrete-time cellular neural networks as a model of excitable media. It can be considered as a CNN solution of a reaction-diffusion equation. This approach adapts the cellular automation of Gerhardt and Schuster (1989) to the CNN paradigm. It is shown that a large variety of complex patterns (including various types of spiral waves) can be efficiently obtained by the proper choice of the model parameters.<>
{"title":"CNN models of complex pattern formation in excitable media","authors":"S. Jankowski, R. Wanczuk","doi":"10.1109/CNNA.1994.381657","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381657","url":null,"abstract":"The paper presents the nonlinear discrete-time cellular neural networks as a model of excitable media. It can be considered as a CNN solution of a reaction-diffusion equation. This approach adapts the cellular automation of Gerhardt and Schuster (1989) to the CNN paradigm. It is shown that a large variety of complex patterns (including various types of spiral waves) can be efficiently obtained by the proper choice of the model parameters.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"51 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":"122026257","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381676
J. P. D. Gyvez
The paper presents a software prototype capable of performing image processing applications using cellular neural networks (CNN). The software is based on a CNN multi-layer structure in which each primary color is assigned to a unique layer. This allows an added flexibility as different processing applications can be performed in parallel. To be able to handle a full range of color tones, two novel color mapping schemes were derived. In the proposed schemes the color information is obtained from the cell's state rather than from its output. Additionally, a post processor capable of performing pixelwise logical operations among color layers was developed to enhance the results obtained from CNN.<>
{"title":"XCNN: a software package for color image processing","authors":"J. P. D. Gyvez","doi":"10.1109/CNNA.1994.381676","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381676","url":null,"abstract":"The paper presents a software prototype capable of performing image processing applications using cellular neural networks (CNN). The software is based on a CNN multi-layer structure in which each primary color is assigned to a unique layer. This allows an added flexibility as different processing applications can be performed in parallel. To be able to handle a full range of color tones, two novel color mapping schemes were derived. In the proposed schemes the color information is obtained from the cell's state rather than from its output. Additionally, a post processor capable of performing pixelwise logical operations among color layers was developed to enhance the results obtained from CNN.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"89 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":"123037432","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381646
T. Roska, P. Szolgay, Akos Zadndy, P. L. Venetianer, A. Radványi, Tamas Sziranyi
An analogic CNN chip prototyping and development system was designed and manufactured to test and measure different VLSI implementations of the analogic CNN Universal Machine. A high level language was developed to support the design of analogic algorithms and an image capture was designed for on-chip image sensing and through CCD camera.<>
{"title":"On a CNN chip-prototyping system","authors":"T. Roska, P. Szolgay, Akos Zadndy, P. L. Venetianer, A. Radványi, Tamas Sziranyi","doi":"10.1109/CNNA.1994.381646","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381646","url":null,"abstract":"An analogic CNN chip prototyping and development system was designed and manufactured to test and measure different VLSI implementations of the analogic CNN Universal Machine. A high level language was developed to support the design of analogic algorithms and an image capture was designed for on-chip image sensing and through CCD camera.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"43 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":"123342814","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381691
W. Utschick, J. Nossek
The theory of probably approximately correct (PAC) learning is applied to discrete-time cellular neural networks (DTCNNS). The Vapnik-Chervonenkis dimension of DTCNN is determined. Considering two different operation modes of the network, an upper bound of the sample size for a reliable generalization of DTCNN architecture is given.<>
{"title":"Computational learning theory applied to discrete-time cellular neural networks","authors":"W. Utschick, J. Nossek","doi":"10.1109/CNNA.1994.381691","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381691","url":null,"abstract":"The theory of probably approximately correct (PAC) learning is applied to discrete-time cellular neural networks (DTCNNS). The Vapnik-Chervonenkis dimension of DTCNN is determined. Considering two different operation modes of the network, an upper bound of the sample size for a reliable generalization of DTCNN architecture is given.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"47 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1994-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120927317","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}
Pub Date : 1994-12-18DOI: 10.1109/CNNA.1994.381628
M. Kanaya, M. Tanaka
We propose a novel method based on the local current comparison method for planning the moving paths of a multi-robot and give some simulation results. This method uses an analog resistive network, competitive networks to find the maximum local current, and a digital-type cellular neural network to search the path. The local current comparison method is related to neighbour node analysis, and this method is suitable as the hardware on the analog-digital hybrid chip. Its basic principle is based on analog dynamics, and it makes the plans so fast that plans can be generated in real-time for robots moving comparatively quickly.<>
{"title":"Robot multi-driving controls by cellular neural networks","authors":"M. Kanaya, M. Tanaka","doi":"10.1109/CNNA.1994.381628","DOIUrl":"https://doi.org/10.1109/CNNA.1994.381628","url":null,"abstract":"We propose a novel method based on the local current comparison method for planning the moving paths of a multi-robot and give some simulation results. This method uses an analog resistive network, competitive networks to find the maximum local current, and a digital-type cellular neural network to search the path. The local current comparison method is related to neighbour node analysis, and this method is suitable as the hardware on the analog-digital hybrid chip. Its basic principle is based on analog dynamics, and it makes the plans so fast that plans can be generated in real-time for robots moving comparatively quickly.<<ETX>>","PeriodicalId":248898,"journal":{"name":"Proceedings of the Third IEEE International Workshop on Cellular Neural Networks and their Applications (CNNA-94)","volume":"22 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":"130358355","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：