Subversion Repositories dashGPS

/* ----------------------------------------------------------------------

* Copyright (C) 2010-2018 Arm Limited. All rights reserved.

*

*

* Project:       CMSIS NN Library

* Title:         arm_nnexamples_cifar10.cpp

*

* Description:   Convolutional Neural Network Example

*

* Target Processor: Cortex-M4/Cortex-M7

*

* 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 Arm LIMITED 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 OWNER 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.

* -------------------------------------------------------------------- */

/**

 * @ingroup groupExamples

 */

/**

 * @defgroup CNNExample Convolutional Neural Network Example

 *

 * \par Description:

 * \par

 * Demonstrates a convolutional neural network (CNN) example with the use of convolution,

 * ReLU activation, pooling and fully-connected functions.

 *

 * \par Model definition:

 * \par

 * The CNN used in this example is based on CIFAR-10 example from Caffe [1].

 * The neural network consists

 * of 3 convolution layers interspersed by ReLU activation and max pooling layers, followed by a

 * fully-connected layer at the end. The input to the network is a 32x32 pixel color image, which will

 * be classified into one of the 10 output classes.

 * This example model implementation needs 32.3 KB to store weights, 40 KB for activations and

 * 3.1 KB for storing the \c im2col data.

 *

 * \image html CIFAR10_CNN.gif "Neural Network model definition"

 *

 * \par Variables Description:

 * \par

 * \li \c conv1_wt, \c conv2_wt, \c conv3_wt are convolution layer weight matrices

 * \li \c conv1_bias, \c conv2_bias, \c conv3_bias are convolution layer bias arrays

 * \li \c ip1_wt, ip1_bias point to fully-connected layer weights and biases

 * \li \c input_data points to the input image data

 * \li \c output_data points to the classification output

 * \li \c col_buffer is a buffer to store the \c im2col output

 * \li \c scratch_buffer is used to store the activation data (intermediate layer outputs)

 *

 * \par CMSIS DSP Software Library Functions Used:

 * \par

 * - arm_convolve_HWC_q7_RGB()

 * - arm_convolve_HWC_q7_fast()

 * - arm_relu_q7()

 * - arm_maxpool_q7_HWC()

 * - arm_avepool_q7_HWC()

 * - arm_fully_connected_q7_opt()

 * - arm_fully_connected_q7()

 *

 * <b> Refer  </b>

 * \link arm_nnexamples_cifar10.cpp \endlink

 *

 * \par [1] https://github.com/BVLC/caffe

 */

#include <stdint.h>

#include <stdio.h>

#include "arm_math.h"

#include "arm_nnexamples_cifar10_parameter.h"

#include "arm_nnexamples_cifar10_weights.h"

#include "arm_nnfunctions.h"

#include "arm_nnexamples_cifar10_inputs.h"

#ifdef _RTE_

#include "RTE_Components.h"

#ifdef RTE_Compiler_EventRecorder

#include "EventRecorder.h"

#endif

#endif

// include the input and weights

static q7_t conv1_wt[CONV1_IM_CH * CONV1_KER_DIM * CONV1_KER_DIM * CONV1_OUT_CH] = CONV1_WT;

static q7_t conv1_bias[CONV1_OUT_CH] = CONV1_BIAS;

static q7_t conv2_wt[CONV2_IM_CH * CONV2_KER_DIM * CONV2_KER_DIM * CONV2_OUT_CH] = CONV2_WT;

static q7_t conv2_bias[CONV2_OUT_CH] = CONV2_BIAS;

static q7_t conv3_wt[CONV3_IM_CH * CONV3_KER_DIM * CONV3_KER_DIM * CONV3_OUT_CH] = CONV3_WT;

static q7_t conv3_bias[CONV3_OUT_CH] = CONV3_BIAS;

static q7_t ip1_wt[IP1_DIM * IP1_OUT] = IP1_WT;

static q7_t ip1_bias[IP1_OUT] = IP1_BIAS;

/* Here the image_data should be the raw uint8 type RGB image in [RGB, RGB, RGB ... RGB] format */

uint8_t   image_data[CONV1_IM_CH * CONV1_IM_DIM * CONV1_IM_DIM] = IMG_DATA;

q7_t      output_data[IP1_OUT];

//vector buffer: max(im2col buffer,average pool buffer, fully connected buffer)

q7_t      col_buffer[2 * 5 * 5 * 32 * 2];

q7_t      scratch_buffer[32 * 32 * 10 * 4];

int main()

{

  #ifdef RTE_Compiler_EventRecorder

  EventRecorderInitialize (EventRecordAll, 1);  // initialize and start Event Recorder

  #endif

  printf("start execution\n");

  /* start the execution */

  q7_t     *img_buffer1 = scratch_buffer;

  q7_t     *img_buffer2 = img_buffer1 + 32 * 32 * 32;

  /* input pre-processing */

  int mean_data[3] = INPUT_MEAN_SHIFT;

  unsigned int scale_data[3] = INPUT_RIGHT_SHIFT;

  for (int i=0;i<32*32*3; i+=3) {

    img_buffer2[i] =   (q7_t)__SSAT( ((((int)image_data[i]   - mean_data[0])<<7) + (0x1<<(scale_data[0]-1)))

                             >> scale_data[0], 8);

    img_buffer2[i+1] = (q7_t)__SSAT( ((((int)image_data[i+1] - mean_data[1])<<7) + (0x1<<(scale_data[1]-1)))

                             >> scale_data[1], 8);

    img_buffer2[i+2] = (q7_t)__SSAT( ((((int)image_data[i+2] - mean_data[2])<<7) + (0x1<<(scale_data[2]-1)))

                             >> scale_data[2], 8);

  }

  // conv1 img_buffer2 -> img_buffer1

  arm_convolve_HWC_q7_RGB(img_buffer2, CONV1_IM_DIM, CONV1_IM_CH, conv1_wt, CONV1_OUT_CH, CONV1_KER_DIM, CONV1_PADDING,

                          CONV1_STRIDE, conv1_bias, CONV1_BIAS_LSHIFT, CONV1_OUT_RSHIFT, img_buffer1, CONV1_OUT_DIM,

                          (q15_t *) col_buffer, NULL);

  arm_relu_q7(img_buffer1, CONV1_OUT_DIM * CONV1_OUT_DIM * CONV1_OUT_CH);

  // pool1 img_buffer1 -> img_buffer2

  arm_maxpool_q7_HWC(img_buffer1, CONV1_OUT_DIM, CONV1_OUT_CH, POOL1_KER_DIM,

                     POOL1_PADDING, POOL1_STRIDE, POOL1_OUT_DIM, NULL, img_buffer2);

  // conv2 img_buffer2 -> img_buffer1

  arm_convolve_HWC_q7_fast(img_buffer2, CONV2_IM_DIM, CONV2_IM_CH, conv2_wt, CONV2_OUT_CH, CONV2_KER_DIM,

                           CONV2_PADDING, CONV2_STRIDE, conv2_bias, CONV2_BIAS_LSHIFT, CONV2_OUT_RSHIFT, img_buffer1,

                           CONV2_OUT_DIM, (q15_t *) col_buffer, NULL);

  arm_relu_q7(img_buffer1, CONV2_OUT_DIM * CONV2_OUT_DIM * CONV2_OUT_CH);

  // pool2 img_buffer1 -> img_buffer2

  arm_maxpool_q7_HWC(img_buffer1, CONV2_OUT_DIM, CONV2_OUT_CH, POOL2_KER_DIM,

                     POOL2_PADDING, POOL2_STRIDE, POOL2_OUT_DIM, col_buffer, img_buffer2);

// conv3 img_buffer2 -> img_buffer1

  arm_convolve_HWC_q7_fast(img_buffer2, CONV3_IM_DIM, CONV3_IM_CH, conv3_wt, CONV3_OUT_CH, CONV3_KER_DIM,

                           CONV3_PADDING, CONV3_STRIDE, conv3_bias, CONV3_BIAS_LSHIFT, CONV3_OUT_RSHIFT, img_buffer1,

                           CONV3_OUT_DIM, (q15_t *) col_buffer, NULL);

  arm_relu_q7(img_buffer1, CONV3_OUT_DIM * CONV3_OUT_DIM * CONV3_OUT_CH);

  // pool3 img_buffer-> img_buffer2

  arm_maxpool_q7_HWC(img_buffer1, CONV3_OUT_DIM, CONV3_OUT_CH, POOL3_KER_DIM,

                     POOL3_PADDING, POOL3_STRIDE, POOL3_OUT_DIM, col_buffer, img_buffer2);

  arm_fully_connected_q7_opt(img_buffer2, ip1_wt, IP1_DIM, IP1_OUT, IP1_BIAS_LSHIFT, IP1_OUT_RSHIFT, ip1_bias,

                             output_data, (q15_t *) img_buffer1);

  arm_softmax_q7(output_data, 10, output_data);

  for (int i = 0; i < 10; i++)

  {

      printf("%d: %d\n", i, output_data[i]);

  }

  return 0;

}