Fuzzy reasoning in image processing has been proved to be a very effective way to formalize complex inference techniques based on heuristics or experience, taking perceptual quality criteria into account. In this pape...
详细信息
Fuzzy reasoning in image processing has been proved to be a very effective way to formalize complex inference techniques based on heuristics or experience, taking perceptual quality criteria into account. In this paper, we discuss implementation of fuzzy reasoning image processing on the standard cellular neural network universal machine. In this way, it is possible to employ such powerful massively parallel chips to speed up use of known algorithms, and to systematize design of new perceptual-quality driven CNN applications.
The maximum operating frequency of a priority encoder is usually limited by the long propagation delay of the priority token, and the delay will increase as the number of bit of the priority encoder increases. The con...
详细信息
ISBN:
(纸本)0780354826
The maximum operating frequency of a priority encoder is usually limited by the long propagation delay of the priority token, and the delay will increase as the number of bit of the priority encoder increases. The concept of look-ahead can be applied to improve the performance. In this paper, the design of a high-speed priority encoder is presented. The main idea of this new design is a multi-level look-ahead structure, which can be realized efficiently by the single-phase-clocked dynamic CMOS logic. A 32-bit priority encoder is implemented in a 3 V, 0.6 /spl mu/m CMOS technology to evaluate the performance of proposed techniques. The new priority encoder uses a 3-level look-ahead structure. As compared with the conventional design, the new design achieves 57% speed improvement with 5% layout area reduction.
To overcome the large extra hardware overhead attendant in the full scan design, the concept of partial scan designs has emerged with the virtue of less area and testability close to full scan. In this paper we analyz...
详细信息
To overcome the large extra hardware overhead attendant in the full scan design, the concept of partial scan designs has emerged with the virtue of less area and testability close to full scan. In this paper we analyzed and unified the strength of the techniques by structural analysis and testabilities. The new partial scan design proposed not only reduces the time for selecting scan flip-flops but also preserves high fault coverage. Test results demonstrate the high fault coverage and remarkable reduction in time for the most ISCAS89 benchmark circuits.
This paper describes an application of the Verilog-A, which is a hardware description language for analog applications, to the modeling of neural networks. We attempt to simulate neural networks having a learning algo...
详细信息
This paper describes an application of the Verilog-A, which is a hardware description language for analog applications, to the modeling of neural networks. We attempt to simulate neural networks having a learning algorithm, which has not been designed with electroniccircuits. The learning algorithm is modeled with Verilog-A and the suitable synaptic weights are solved by Verilog-A simulation.
Direct tuning techniques of continuous-time filters for mobile radio receivers are discussed. Different design aspects and the effect of nonidealities are considered. A single-shot tuning scheme which is based on digi...
详细信息
Direct tuning techniques of continuous-time filters for mobile radio receivers are discussed. Different design aspects and the effect of nonidealities are considered. A single-shot tuning scheme which is based on digital control word calculation algorithm is proposed for cases (CDMA) where sufficient time cannot be allocated for feedback based tuning as in time-division multiplexed radio systems (TDMA).
An approach is presented for the automated sizing of analog circuits based upon a combination of a genetic algorithm (GA) with a least squares (Gauss-Newton) gradient search. The method combines the global-search prop...
详细信息
An approach is presented for the automated sizing of analog circuits based upon a combination of a genetic algorithm (GA) with a least squares (Gauss-Newton) gradient search. The method combines the global-search properties of the GA with the fast local convergence properties of the least squares method to produce a circuit design from random initial component values in a reduced time compared to the application of a direct GA method, or a restart least squares algorithm. Results are presented to demonstrate the application of the method in the design of both passive and active circuits.
The truncation of binary numbers can be modeled as noise and a DC-error if certain conditions are fulfilled. These truncation effects in finite impulse response (FIR) digital filters that are realized with canonic sig...
详细信息
The truncation of binary numbers can be modeled as noise and a DC-error if certain conditions are fulfilled. These truncation effects in finite impulse response (FIR) digital filters that are realized with canonic signed-digit (CSD) representation are reviewed. Three different techniques to reduce or eliminate the DC-error are presented and compared with respect to hardware and power consumption requirements.
This paper describes a methodology for designing analog systems starting from a top-level, behavioral model. The methodology has been incorporated within an analog synthesis system (ADSA), which is based upon the use ...
详细信息
This paper describes a methodology for designing analog systems starting from a top-level, behavioral model. The methodology has been incorporated within an analog synthesis system (ADSA), which is based upon the use of optimization at both behavioral and circuit levels. Initial design uses ideal models without interaction. Model refinement takes place to accommodate, first, signal offsets and levels, and then, the source and load impedances of the circuit blocks. An example is given of the methodology applied to the design of a Downconverter.
This paper presents a new effective method for symbolic sensitivity analysis of large scale analogue circuits. Based on the "sequence of expressions" (SOE) approach the sensitivities with respect to all para...
详细信息
This paper presents a new effective method for symbolic sensitivity analysis of large scale analogue circuits. Based on the "sequence of expressions" (SOE) approach the sensitivities with respect to all parameters are calculated in parallel. Experimental results show that a significant acceleration compared to previously described symbolic procedures is achieved and that the method can be faster than the numerical adjoint approach.
Genetic algorithms are applied to design multilayer discrete-time cellular neural networks for image processing tasks, To this end not only the templates of the different layers will be optimized, but also the network...
详细信息
Genetic algorithms are applied to design multilayer discrete-time cellular neural networks for image processing tasks, To this end not only the templates of the different layers will be optimized, but also the network structure itself, that is, number of layers and iterations per layer. As a difference with traditional strategies, both the definition of the optimum network size and the template optimization are done simultaneously.
暂无评论