深度学习之卷积神经网络CNN模式识别VS代码

深度学习之卷积神经网络CNN做手写体识别的VS代码。支持linux版本和VS2012版本。 tiny-cnn: A C++11 implementation of convolutional neural networks ======== tiny-cnn is a C++11 implementation of convolutional neural networks. design principle ----- * fast, without GPU 98.8% accuracy on MNIST in 13 minutes training (@Core i7-3520M) * header only, policy-based design supported networks ----- ### layer-types * fully-connected layer * convolutional layer * average pooling layer ### activation functions * tanh * sigmoid * rectified linear * identity ### loss functions * cross-entropy * mean-squared-error ### optimization algorithm * stochastic gradient descent (with/without L2 normalization) * stochastic gradient levenberg marquardt dependencies ----- * boost C++ library * Intel TBB sample code ------ ```cpp #include "tiny_cnn.h" using namespace tiny_cnn; // specify loss-function and optimization-algorithm typedef network CNN; // tanh, 32x32 input, 5x5 window, 1-6 feature-maps convolution convolutional_layer C1(32, 32, 5, 1, 6); // tanh, 28x28 input, 6 feature-maps, 2x2 subsampling average_pooling_layer S2(28, 28, 6, 2); // fully-connected layers fully_connected_layer F3(14*14*6, 120); fully_connected_layer F4(120, 10); // connect all CNN mynet; mynet.add(&C1); mynet.add(&S2); mynet.add(&F3); mynet.add(&F4); assert(mynet.in_dim() == 32*32); assert(mynet.out_dim() == 10); ``` more sample, read main.cpp build sample program ------ ### gcc(4.6~) without tbb ./waf configure --BOOST_ROOT=your-boost-root ./waf build with tbb ./waf configure --TBB --TBB_ROOT=your-tbb-root --BOOST_ROOT=your-boost-root ./waf build with tbb and SSE/AVX ./waf configure --AVX --TBB --TBB_ROOT=your-tbb-root --BOOST_ROOT=your-boost-root ./waf build ./waf configure --SSE --TBB --TBB_ROOT=your-tbb-root --BOOST_ROOT=your-boost-root ./waf build or edit inlude/co

/*
    Copyright （c） 2013 Taiga Nomi
    All rights reserved.
    
    Redistribution and use in source and binary forms with or without
    modification are permitted provided that the following conditions are met:
    * Redistributions of source code must retain the above copyright
    notice this list of conditions and the following disclaimer.
    * Redistributions in binary form must reproduce the above copyright
    notice this list of conditions and the following disclaimer in the
    documentation and/or other materials provided with the distribution.
    * Neither the name of the  nor the
    names of its contributors may be used to endorse or promote products
    derived from this software without specific prior written permission.

    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS“ AND ANY 
    EXPRESS OR IMPLIED WARRANTIES INCLUDING BUT NOT LIMITED TO THE IMPLIED 
    WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 
    DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY 
    DIRECT INDIRECT INCIDENTAL SPECIAL EXEMPLARY OR CONSEQUENTIAL DAMAGES 
    （INCLUDING BUT NOT LIMITED TO PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 
    LOSS OF USE DATA OR PROFITS; OR BUSINESS INTERRUPTION） HOWEVER CAUSED AND 
    ON ANY THEORY OF LIABILITY WHETHER IN CONTRACT STRICT LIABILITY OR TORT 
    （INCLUDING NEGLIGENCE OR OTHERWISE） ARISING IN ANY WAY OUT OF THE USE OF THIS 
    SOFTWARE EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#include 
#include 
#include 

#include “tiny_cnn.h“
//#define NOMINMAX
//#include “imdebug.h“

void sample1_3layerNN（）;

using namespace tiny_cnn;

int main（void） {
    // construct LeNet-5 architecture
    typedef network CNN;
    CNN nn;
    convolutional_layer C1（32 32 5 1 6）;
    average_pooling_layer S2（28 28 6 2）;
    // connection table [Y.Lecun 1998 Table.1]
#define O true
#define X false
    static const bool connection[] = {
        O X X X O O O X X O O O O X O O
        O O X X X O O O X X O O O O X O
        O O O X X X O O O X X O X O O O
        X O O O X X O O O O X X O X O O
        X X O O O X X O O O O X O O X O
        X X X O O O X X O O O O X O O O
    };
#undef O
#undef X
    convolutional_layer C3（14 14 5 6 16 connection_table（connection 6 16））;
    average_pooling_layer S4（10 10 16 2）;
    convolutional_layer C5（5 5 5 16 120）;
    fully_connected_layer F6（120 10）;

    assert（C1.param_size（） == 156 && C1.connection_size（） == 122304）;
    assert（S2.param_size（） == 12 && S2.connection_size（） == 5880）;
    assert（C3.param_size（） == 1516 && C

 属性            大小     日期    时间   名称
----------- ---------  ---------- -----  ----

    .......       483  2014-02-11 06:00  cnn_vs2012\.gitattributes

    .......      2107  2014-02-11 06:00  cnn_vs2012\.gitignore

    .......      2053  2014-02-11 06:00  cnn_vs2012\README.md

    .......      3628  2014-02-11 06:00  cnn_vs2012\include\activation_function.h

    .......      3501  2014-02-11 06:00  cnn_vs2012\include\average_pooling_layer.h

    .......      1920  2014-02-11 06:00  cnn_vs2012\include\config.h

    .......      6262  2014-02-11 06:00  cnn_vs2012\include\convolutional_layer.h

    .......      1966  2014-02-11 06:00  cnn_vs2012\include\deform.h

    .......      5793  2014-02-11 06:00  cnn_vs2012\include\fixed_point.h

    .......      5041  2014-02-11 06:00  cnn_vs2012\include\fully_connected_layer.h

    .......      4636  2014-02-11 06:00  cnn_vs2012\include\image.h

    .......     10225  2014-02-11 06:00  cnn_vs2012\include\layer.h

    .......      2097  2014-02-11 06:00  cnn_vs2012\include\loss_function.h

    .......      4612  2014-02-11 06:00  cnn_vs2012\include\mnist_parser.h

    .......     13429  2014-02-11 06:00  cnn_vs2012\include\network.h

    .......      2731  2014-02-11 06:00  cnn_vs2012\include\optimizer.h

    .......      8703  2014-02-11 06:00  cnn_vs2012\include\partial_connected_layer.h

    .......     23363  2014-02-11 06:00  cnn_vs2012\include\picotest.h

    .......     12527  2014-02-11 06:00  cnn_vs2012\include\product.h

    .......      1910  2014-02-11 06:00  cnn_vs2012\include\tiny_cnn.h

    .......      5446  2014-02-11 06:00  cnn_vs2012\include\util.h

    .......      6757  2014-02-11 06:00  cnn_vs2012\src\main.cpp

    .......     15168  2014-02-11 06:00  cnn_vs2012\src\test.cpp

    .......       352  2014-02-11 06:00  cnn_vs2012\src\wscript

    .......      2039  2014-02-11 06:00  cnn_vs2012\vc\cnn.vcxproj.filters

    .......   7840016  2014-02-11 06:00  cnn_vs2012\vc\t10k-images.idx3-ubyte

    .......     10008  2014-02-11 06:00  cnn_vs2012\vc\t10k-labels.idx1-ubyte

    .......      1936  2014-02-11 06:00  cnn_vs2012\vc\test.vcxproj.filters

    .......      1320  2014-02-11 06:00  cnn_vs2012\vc\tiny_cnn.sln

    .......      5856  2014-02-11 06:00  cnn_vs2012\vc\tiny_cnn.vcxproj

............此处省略12个文件信息