CNN Universal Machine Research Articles

A CNN algorithm for real time analysis of DNA microarrays

The work focuses on a new approach for DNA-chip image processing by using the cellular neural network (CNN) paradigm. In particular, the work aims to solve the main issues in automatic classification and spot validation that arise during an automatic DNA microarray analysis procedure. This contribution briefly presents the basic DNA micro-array technique and the state-of-the-art technologies, and introduces the idea of processing DNA chip data via CNNs, reporting suitable examples. The algorithm described has been developed using a high level language, dedicated to CNN universal machine (CNNUM) chip programming, and validated on a typical fluorescence image from a DNA microarray after hybridization.

THE SPATIOTEMPORAL PRISONER'S DILEMMA

A novel platform, the cellular neural network universal machine (CNN-UM) model is proposed to simulate the spatial prisoner's dilemma (SPD) game introduced in [Nowak & May, 1992]. The application of CNN is motivated by both its high conformity with the modeled natural phenomena (analog dynamics) and its superior speed enabling the exploration of the SPD's vast parameter space. Several models are shown as examples of CNN implementation. In particular, a synchronously updated model is introduced using strategy based on best individual strategy (BIS). A new synchronous model is shown based on better average strategy (BAS). Stemming from the CNN implementation of synchronous SPD an asynchronous model has been developed employing BIS, and the survival of clusters of cooperating individuals is demonstrated. This seems to contradict the view elaborated in [Huberman & Glance 1993] claiming that the viability of cooperation is an artifact of synchronous updating in SPD.

Image Segmentation Using Markov Random Field Model in Fully Parallel Cellular Network Architectures

Markovian approaches to early vision processes need a huge amount of computing power. These algorithms can usually be implemented on parallel computing structures. Herein, we show that the Markovian labeling approach can be implemented in fully parallel cellular network architectures, using simple functions and data representations. This makes possible to implement our model in parallel imaging VLSI chips.As an example, we have developed a simplified statistical image segmentation algorithm for the Cellular Neural/Nonlinear Networks Universal Machine (CNN-UM), which is a new image processing tool, containing thousands of cells with analog dynamics, local memories and processing units. The Modified Metropolis Dynamics (MMD) optimization method can be implemented into the raw analog architecture of the CNN-UM. We can introduce the whole pseudo-stochastic segmentation process in the CNN architecture using 8 memories/cell. We use simple arithmetic functions (addition, multiplication), equality-test between neighboring pixels and very simple nonlinear output functions (step, jigsaw). With this architecture, the proposed VLSI CNN chip can execute a pseudo-stochastic relaxation algorithm of about 100 iterations in about 100 μs.In the suggested solution the segmentation is unsupervised, where a pixel-level statistical estimation model is used. We have tested different monogrid and multigrid architectures.In our CNN-UM model several complex preprocessing steps can be involved, such as texture-classification or anisotropic diffusion. With these preprocessing steps, our fully parallel cellular system may work as a high-level image segmentation machine, using only simple functions based on the close-neighborhood of a pixel.

An 0.5-μm CMOS analog random access memory chip for TeraOPS speed multimedia video processing

Data compressing, data coding, and communications in object-oriented multimedia applications like telepresence, computer-aided medical diagnosis, or telesurgery require an enormous computing power-in the order of trillions of operations per second (TeraOPS). Compared with conventional digital technology, cellular neural/nonlinear network (CNN)-based computing is capable of realizing these TeraOPS-range image processing tasks in a cost-effective implementation. To exploit the computing power of the CNN Universal Machine (CNN-UM), the CNN chipset architecture has been developed-a mixed-signal hardware platform for CNN-based image processing. One of the nonstandard components of the chipset is the cache memory of the analog array processor, the analog random access memory (ARAM). This paper reports on an ARAM chip that has been designed and fabricated in a 0.5-/spl mu/m CMOS technology. This chip consists of a fully addressable array of 32/spl times/256 analog memory registers and has a packing density of 637 analog-memory-cells/mm/sup 2/. Random and nondestructive access of the memory contents is available. Bottom-plate sampling techniques have been employed to eliminate harmonic distortion introduced by signal-dependent feedthrough. Signal coupling and interaction have been minimized by proper layout measures, including the use of protection rings and separate power supplies for the analog and the digital circuitry. This prototype features an equivalent resolution of up to 7 bits-measured by comparing the reconstructed waveform with the original input signal. Measured access times for writing/reading to/from the memory registers are of 200 ns. I/O rates via the l6-line-wide I/O bus exceed 10 Msamples/s. Storage time at room temperature is in the 80 to 100 ms range, without accuracy loss.

Fault-tolerant design of analogic CNN templates and algorithms-Part I: The binary output case

This paper addresses the issue of designing a class of fault-tolerant cellular neural network (CNN) templates that, combined with CNN analogic algorithms, work correctly and reliably on given CNN universal machine (CNN-UM) chips. In particular, a generic method for finding nonpropagating binary-output CNN templates is proposed. This method is based on measurements of actual CNN-UM chips and combines adaptive optimization and decomposition of theoretically ideal CNN templates in order to correct the erroneous behavior of actual CNN-UM chips, which is mainly caused by imperfections introduced during fabrication. More specifically, the entire array of cells in a CNN-UM chip is modeled by a single feed-forward virtual cell whose optimal parameters are found by a simple and effective gradient-based method. In the case of binary input-output uncoupled templates (or Boolean operators), a systematic template decomposition method is applied whenever optimization fails to find a correctly working CNN template for all possible combinations of local 3/spl times/3 binary input patterns. The resulting templates are finally combined, yielding a simple CNN analogic algorithm. Examples are presented for both binary- and analog-input operators, using two concrete stored-program CNN-UM chips to demonstrate the effectiveness of the proposed method, whose advantages and limitations are also discussed.

Very low bit-rate video coding using cellular neural network universal machine

A method of video coding for very low bit-rate channels, which is implemented using cellular neural network universal machine, is presented in the following paper. The presented method combines elements of a standard approach to video coding with elements of second-generation video-coding techniques. Inter-frame coding is performed using standard block-based motion estimation procedure, while intra-frame coding is based on vector quantization approach. To satisfy constraints imposed by very low bit-rate channel throughput, a number of bytes that were considered to be used for representing video sequence frames, was assumed to be less than 200. Simulations of the algorithm execution, based on actual CNN UM chip parameter values, show feasibility of using the proposed method for real-time implementation of very low bit-rate video coding.

A QCIF Resolution Binary I/O CNN-UM Chip

In this paper is reported a Cellular Nonlinear Network Universal Machine realization where there are 176 × 144 active cells. The size of the network is the standardized QCIF video image format and the design is aimed to be used in segmenting video images in future video applications requiring very low bit-rate for image transmission. The achieved cell density is 3000 cells/mm^2 with a 0.25 micron standard digital CMOS process. Different building blocks inside the cell are given in detail and also the other implemented circuitry is thoroughly discussed. The physical realization of the design is also reported.

Analogic algorithms for optical detection of breaks and short circuits on the layouts of printed circuit boards using CNN

Printed circuit board layout inspection methods are mostly based on local geometric information, therefore they are well suited to the CNN paradigm. Here the detection of two layout errors is considered namely, the breaks in the wires and some kind of short circuits. The designed analogic algorithms to solve the problems above were tested on real-life examples using an experimental system based on our CNN-HAB digital multiprocessor add-in-board with 1 million cell space and at a speed of 2.0 μs/cell/iteration. The break detection algorithm was tested on a 20*22 CNN Universal Machine (CNNUM) chip and the results were compared to other automatic optical inspection systems as well.

An Emulated Digital CNN Implementation

A new emulated digital CNN Universal Machine chip architecture is introduced and the main steps of the design process are shown in this paper. One core processor can be implemented on 2 × 2 mm^2 silicon area with a 0.35 μm CMOS technology. Assuming an array of 24 processors on a chip, its speed is 1ns/virtual cell/CNN iteration with 12 bit precision. This enables the execution of over five hundred 3 × 3 convolution operations on each frame of a 240 × 320-pixel 25 fps digital image flow. Another new feature of the design is its variable precision capability. This allows the user to trade off precision for speed. The architecture supports some non-linear filter implementation as well.

Implementation of large-neighbourhood non-linear templates on the CNN universal machine

In this paper we present a general method for implementing certain types of arbitrary large neighbourhood non-linear control (B) templates on the cellular neural network universal machine (CNNUM). First, we show how a 3×3 non-linear B template can be replaced by a set of linear ones. We thereby obtain an analogic algorithm that, by using linear templates, yields the same result as when applying the original non-linear template. It is then shown how the template decomposition problem can be solved for large neighbourhood non-linear B templates. Thirdly, the question of template decomposition is discussed for arbitrary templates in case of discrete time CNN (DTCNN). Examples are given to highlight the main steps of the algorithm. The necessary and sufficient conditions for realizing the proposed method on any CNNUM chip are also specified. An idea for solving the non-linear feedback (A) template decomposition problem is also proposed. © 1998 John Wiley & Sons, Ltd.

Invertible operations on a Cellular Neural Network Universal Machine

This paper is concerned with the invertibility of parallel operations definable on two-dimensional binary arrays by means of a local Boolean function. In this way, it contributes to the theory of cellular automata and may open new potentialities to cellular neural network applications. We provide a theoretically new approach to the invertibility of two-dimensional additive maps which can be used to obtain detailed information on the behaviour of these operations. Other ways to make invertible operations are also investigated, including local maps which are additive in only one of the terms, and second-order cellular automata. The discussed techniques yield a rich variety of invertible operations on infinite and/or finite two-dimensional binary arrays. When implemented on CNN Universal Chips, all of these operations, performed by TeraOPS speed, become subroutines of a 2D analogic algorithm. These operations can be used in 2D cryptography. © 1998 John Wiley & Sons, Ltd.

Implementation of arbitrary Boolean functions on a CNN Universal Machine

The demand for implementing arbitrary N-variable logic functions on perceptron-like structures arises quite often in practice. It is well known, that only the linearly separable class of Boolean functions can be implemented in a single step on these structures. This class however, constitutes only a small (and with the increasing N exponentially decreasing) part of all logic functions. All Boolean functions can be trivially decomposed into at most 2N−1 sub-functions by implementing the minterm or maxterm formulation of the function. This approach, however, becomes rather impractical with the increasing number of variables. In this paper an algorithm is proposed for decomposing arbitrary, linearly non-separable Boolean functions into a series of separable functions, which can be then efficiently implemented as a program for the CNN Universal Machine assuming the simplest and most robust hardware architecture. The decomposition is achieved by finding the closest linearly separable compact functions. Robustness issues of the implementation are also addressed. © 1998 John Wiley & Sons, Ltd.

CNN‐based difference‐controlled adaptive non‐linear image filters

In this paper, we develop a common cellular neural network framework for various adaptive non-linear filters based on robust statistic and geometry-driven diffusion paradigms. The base models of both approaches are defined as difference-controlled non-linear CNN templates, while the self-adjusting property is ensured by simple analogic (analog and logic) CNN algorithms. Two adaptive strategies are shown for the order statistic class. When applied to the images distorted by impulse noise both give more visually pleasing results with lower-frequency weighted mean square error than the median base model. Generalizing a variational approach we derive the constrained anisotropic diffusion, where the output of the geometry-driven diffusion model is forced to stay close to a pre-defined morphological constraint. We propose a coarse-grid CNN approach that is capable of calculating an acceptable noise-level estimate (proportional to the variance of the Gaussian noise) and controlling the fine-grid anisotropic diffusion models. A combined geometrical-statistical approach has also been developed for filtering both the impulse and additive Gaussian noise while preserving the image structure. We briefly discuss how these methods can be embedded into a more complex algorithm performing edge detection and image segmentation. The design strategies are analysed primarily from VLSI implementation point of view; therefore all non-linear cell interactions of the CNN architecture are reduced to two fundamental non-linearities, to a sigmoid type and a radial basis function. The proposed non-linear characteristics can be approximated with simple piecewise-linear functions of the voltage difference of neighbouring cells. The simplification makes it possible to convert all space-invariant non-linear templates of this study to a standard instruction set of the CNN Universal Machine, where each instruction is coded by at most a dozen analog numbers. Examples and simulation results are given throughout the text using various intensity images.

Image compression by cellular neural networks

This paper demonstrates how the cellular neural-network universal machine (CNNUM) architecture can be applied to image compression. We present a spatial subband image-compression method well suited to the local nature of the CNNUM. In case of lossless image compression, it outperforms the JPEG image-compression standard both in terms of compression efficiency and speed. It performs especially well with radiographical images (mammograms); therefore, it is suggested to use it as part of a cellular neural/nonlinear (CNN)-based mammogram-analysis system. This paper also gives a CNN-based method for the fast implementation of the moving pictures experts group (MPEG) and joint photographic experts group (JPEG) moving and still image-compression standards.

Implementation of binary and gray-scale mathematical morphology on the CNN universal machine

A cellular neural network(CNN)-based morphological engine is proposed. An effective implementation method of binary and gray-scale erosion, dilation, and reconstruction is introduced. The binary morphological operators are successfully implemented on an actual CNN universal chip. Experimental results are shown.

CNN: A Vision of Complexity

CNN: A Vision of Complexity

Object-oriented image analysis for very-low-bitrate video-coding systems using the CNN universal machine

The CNN universal machine (CNNUM) is applied to object-oriented video compression and proves its universality for future applications in the field of very-low-bitrate coding. This proposal joins recent work of Venetianer and Roska in unfolding the enormous computational abilities of the CNNUM for a wide class of video compression techniques. Here a novel image analysis technique is considered and realized in the form of analogic CNN algorithms. The specific features of the scheme, among them the extensive use of dynamic (finite running time) CNN cloning templates, are outlined and discussed through different computer simulations. When implemented on the CNNUM, its performances outdo those of equivalent digital systems and qualify the CNNUM as a serious competitor for future video coding hardware. © 1997 John Wiley & Sons, Ltd.

Object‐oriented image analysis for very‐low‐bitrate video‐coding systems using the CNN universal machine

The CNN universal machine (CNNUM) is applied to object-oriented video compression and proves its universality for future applications in the field of very-low-bitrate coding. This proposal joins recent work of Venetianer and Roska in unfolding the enormous computational abilities of the CNNUM for a wide class of video compression techniques. Here a novel image analysis technique is considered and realized in the form of analogic CNN algorithms. The specific features of the scheme, among them the extensive use of dynamic (finite running time) CNN cloning templates, are outlined and discussed through different computer simulations. When implemented on the CNNUM, its performances outdo those of equivalent digital systems and qualify the CNNUM as a serious competitor for future video coding hardware. © 1997 John Wiley & Sons, Ltd.

A 0.8-μm CMOS two-dimensional programmable mixed-signal focal-plane array processor with on-chip binary imaging and instructions storage

This paper presents a CMOS chip for the parallel acquisition and concurrent analog processing of two-dimensional (2-D) binary images. Its processing function is determined by a reduced set of 19 analog coefficients whose values are programmable with 7-b accuracy. The internal programming signals are analog, but the external control interface is fully digital. On-chip nonlinear digital-to-analog converters (DAC's) map digitally coded weight values into analog control signals, using feedback to predistort their transfer characteristics in accordance to the response of the analog programming circuitry. This strategy cancels out the nonlinear dependence of the analog circuitry with the programming signal and reduces the influence of interchip technological parameters random fluctuations. The chip includes a small digital RAM memory to store eight sets of processing parameters in the periphery of the cell array and four 2-D binary images spatially distributed over the processing array. It also includes the necessary control circuitry to realize the stored instructions in any order and also to realize programmable logic operations among images. The chip architecture is based on the cellular neural/nonlinear network universal machine (CNN-UM). It has been fabricated in a 0.8-/spl mu/m single-poly double-metal technology and features 2-/spl mu/s operation speed (time required to process an image) and around 7-b accuracy in the analog processing operations.

Efficient implementation of neighborhood logic for cellular automata via the Cellular Neural Network Universal Machine

The main difficulty in implementing cellular automata on the Cellular Neural Network Universal Machine (CNNUM) is the need to perform arbitrary logic functions of the input neighborhood. Since the architecture computes weighted sums of this neighborhood, by using a B-template, it is limited to threshold logic, i.e., a logical operation to be computed by a single transient must be in the class of linearly separable Boolean functions. It was shown previously how a general logic function can be implemented on the CNNUM by cascading component functions from this class-namely by the direct implementation of the minterm or maxterm formulation of the desired function. However, for functions of a 3/spl times/3 input neighborhood this method may require up to 256 stages. We propose a more efficient method for implementing general logic functions on the CNNUM and other hardwares capable of performing a threshold logic function of an input neighborhood. The class of considered component functions is a superset of the minterms and maxterms but, for purposes of searchability, ease of implementation, and robustness, a subset of the general linearly separable Boolean functions. We have formulated an algorithm that will find a sequence of weight-restricted threshold logic functions (B-templates with weights from {-1, 0, +1} and a bias) that, when cascaded together using two-input logical operations, will result in the desired Boolean function. Two examples are given to exhibit the algorithm.