卷積神經(jīng)網(wǎng)絡(luò)(參考版)
2024-10-03 00:36本頁面
【正文】 Shape recognition內(nèi)容總結(jié)
（1）卷積神經(jīng)網(wǎng)絡(luò)
摘要：卷積神經(jīng)網(wǎng)絡(luò)是近年來廣泛應(yīng)用于模式識(shí)別、圖像處理等領(lǐng)域的一種高效識(shí)別算法，它具有結(jié)構(gòu)簡(jiǎn)單、訓(xùn)練參數(shù)少和適應(yīng)性強(qiáng)等特點(diǎn)
（2）考慮到本文使用卷積神經(jīng)網(wǎng)絡(luò)是用作分類器，其類別數(shù)為2（即人臉和非人臉），所以輸出層的節(jié)點(diǎn)數(shù)為2
（3）設(shè)計(jì)方法如下：
卷積層的設(shè)計(jì)：信號(hào)的卷積運(yùn)算是信號(hào)處理領(lǐng)域中最重要的運(yùn)算之一
。 Advantage。參考文獻(xiàn)[1] 王天翼. 基于卷積網(wǎng)絡(luò)的三位特征提取 [學(xué)位論文].吉林:吉林大學(xué),2006.[2] 李葆青. 基于卷積神經(jīng)網(wǎng)絡(luò)的模式分類器 [J].大連大學(xué)學(xué)報(bào),2003,24(2):1923.[3] Simon Haykin 著,葉世偉, [M]. 北京：機(jī)械工業(yè)出版社，2004.[4] [學(xué)位論文].北京:華北電力大學(xué), 2009.[5] 陸璐. 卷積神經(jīng)網(wǎng)絡(luò)的研究及其在車牌識(shí)別系統(tǒng)中的應(yīng)用 [學(xué)位論文].合肥:合肥工業(yè)大學(xué),2006.[6] 顧佳玲, [J].系統(tǒng)仿真學(xué)報(bào), 2009,21(8):24412445.[7] 趙志宏,楊紹普, [J].系統(tǒng)仿真學(xué)報(bào),2010,22(3):638641.[8] T .H .Hildebrandt, Optimal Training of Thresholded Linear Correlation Classifiers, IEEE Transactions on Neural Networks Z(6),Nov.(1991).[9] K. Fukushima，“Neocognitron:A selforganizing neuralnetwork model for a Mechanism of Pattern recognition unaffected by shift in position，Biol. Cybern.，1980.[10] ，Shape, position and size invariant visual pattern recognition based on principles of neocognitron and perception in Artificial Neural Networks，Eds. Amsterdam the Netherlands: North一Holland，1992，.[11] D. Lovell，et al.，Comments on“Optimal Training of Thresholded Linear Correlation Classifiers”,IEEE Trans .On Neural Networks 4(2),March(1993). [12] K. Fukushima，“Analysis of the process of visual pattern recognition by the neocognitron，”Neural Networks，1989 [13] Van Ooyen and B. Nienhuis，Pattern Recognition in the Neocognitron Is Improved一by Neuronal Adaption，Biological Cyberneties70,(1993).[14] 石大明，劉海濤，[J」，2001，24(5):468一473[15] Claus Neubauer. Evaluation of’ Convolutional Neural Networks for VisualRecognition，Neural Netwoks，(1998)[16] and ,“Use of different thresholds in learning and recognition，”NeuroeomPuting，1996.[17] G. W. Cottrell，“EMPATH: Face，emotion，and gender recognition using holons，”in Advances in Neural Information Processing Sys[18] H. Bourlard and Y. Kamp，“Autoassoziation by multilayerperceptrons and singular value deposition，”Biol. Cybern.，1988.[19]洪家榮，李星原. Neocognitron學(xué)習(xí)算法分析. 軟件學(xué)報(bào) [J]，1994，5，(4):3539[20]Fukushima K. A hierarchical neural network capable of visual pattern recognition . Neural Networks，1989:2:413420[21] ，and ，“Globally Trained Handwritten Word Recognizer using Spatial Representation, Convolutional Neural Networks and Hidden Markov Models in Advances in Neural Information Processing Systems，Jack ，Gerald Tesauro，and Joshua AlsPector，Morgan Kaufmann Publishers，Inc.[22]張佳康,[J],2010,36(51):179181.[23]肖柏旭,[J],2006,26:4648.[24][學(xué)位論文].廈門:廈門大學(xué),2008.Convolution Neural NetworkAbstract:Convolution neural network is an efficient recognition algorithm which is widely used in pattern recognition, image processing and other fields recent has a simple structure, few training parameters and good adaptability and other advantages. In this paper, begin with the history of convolutional neural networks,describes the structure of convolutional neural network,neuron models and training algorithms in d
醫(yī)療健康相關(guān)推薦
鄂ICP備17016276號(hào)-1