An Optical Thresholding Perceptron an Optical Thresholding Perceptron
نویسندگان
چکیده
An implementation of an optical perceptron with a soft optical threshold trained with an adapted BP algorithm is described as a precursor to an optical multilayer perceptron (MLP). It has 64 inputs and ten outputs. The soft threshold is implemented by a liquid crystal light valve. Experimental results on perceptron recall are also reported. The eeect of a modiied grey-scale to weight mapping for weight levels implemented by LCTVs is evaluated based on the results of handwritten digit recognition.
منابع مشابه
A Method for All - Positive Optical Multilayerperceptronsi
The most promising approaches for optical neural networks are based on intensity encoding. However, a serious drawback of intensity encoding is the lack of negative values and optical subtraction, which are essential for rendering neural networks useful. To overcome the need for optical subtraction, a novel training method is described here that is especially useful for optical multilayer neura...
متن کاملCommon Spatial Pattern for Human Identification Based on Finger Vein Images in Radon space
Finger vein is one of the most fitting biometric for identifying individuals. In this paper a new method for finger vein recognition is proposed. First the veins are extracted from finger vein images by using entropy based thresholding. In finger vein images the veins are appeared as dark lines. The method extracts veins as well, but the images are noisy, that means in addition to the veins the...
متن کاملPublished in the Proceedings of the Third Ieee International Conference on Electronics Circuits and Systems Icecs Rodos Greece October a Method for All Positive Optical Multilayer Perceptrons
The most promising approaches for optical neural networks are based on intensity encoding However a serious drawback of intensity encoding is the lack of negative values and optical subtraction which are essential for rendering neural networks useful To overcome the need for optical subtraction a novel training method is described here that is es pecially useful for optical multilayer neural ne...
متن کاملWavelet based denoising integrated into multilayered perceptron
A denoising unit based on wavelet multiresolution analysis is added ahead of the multilayered perceptron. The cost function used in neural network learning is also applied as the denoising criterion and hence denoising itself is treated as a part of the integrated model. By introducing continuously derivable generalized soft thresholding function and infinite thresholds, a gradient based learni...
متن کاملAn Adaptive Neuro-fuzzy System for Automatic Image Segmentation and Edge Detection
An autoadaptive neuro-fuzzy segmentation and edge detection architecture is presented. The system consist of a multilayer perceptron (MLP)-like network that performs image segmentation by adaptive thresholding of the input image using labels automatically pre-selected by a fuzzy clustering technique. The proposed architecture is feedforward, but unlike the conventional MLP the learning is unsup...
متن کامل