21.Intel二代神经棒使用教程

一、环境配置
1、下载OPENVINO—TOOLKIT
https: // software.intel.com / en - us / openvino - toolkit / choose - download
2、安装OPENVINO
按照默认的路径安装，全部勾选，直接下一步到完成
3、安装CMake，并添加到环境变量
4、安装PYthon，并添加到环境变量
5、设置环境
1) cd
C:\Program
Files(x86)\IntelSWTools\openvino\bin \
    2) setupvars.bat
6、安装对应框架的优化
1）cd
C:\Program
Files(x86)\IntelSWTools\openvino\deployment_tools\model_optimizer\install_prerequisites
2）install_prerequisites_tf.bat  （由于使用的是TensorFlow ）
7、测试
1）cd
C:\Program
Files(x86)\IntelSWTools\openvino\deployment_tools\demo \
    2）demo_squeezenet_download_convert_run.bat （不报错就是安装成功了）

二、模型转换
1、训练模型（以手写体识别为例）
原始TensorFlow手写体识别程序

"""A deep MNIST classifier using convolutional layers.
See extensive documentation at
https://www.tensorflow.org/get_started/mnist/pros
"""
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order

from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import argparse
import sys
import tempfile

from tensorflow.examples.tutorials.mnist import input_data

import tensorflow as tf

FLAGS = None


def deepnn(x):
    """deepnn builds the graph for a deep net for classifying digits.

    Args:
      x: an input tensor with the dimensions (N_examples, 784), where 784 is the
      number of pixels in a standard MNIST image.

    Returns:
      A tuple (y, keep_prob). y is a tensor of shape (N_examples, 10), with values
      equal to the logits of classifying the digit into one of 10 classes (the
      digits 0-9). keep_prob is a scalar placeholder for the probability of
      dropout.
    """
    # Reshape to use within a convolutional neural net.
    # Last dimension is for "features" - there is only one here, since images are
    # grayscale -- it would be 3 for an RGB image, 4 for RGBA, etc.
    with tf.name_scope('reshape'):
        x_image = tf.reshape(x, [-1, 28, 28, 1])

    # First convolutional layer - maps one grayscale image to 32 feature maps.
    with tf.name_scope('conv1'):
        W_conv1 = weight_variable([5, 5, 1, 32])
        b_conv1 = bias_variable([32])
        h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)

    # Pooling layer - downsamples by 2X.
    with tf.name_scope('pool1'):
        h_pool1 = max_pool_2x2(h_conv1)

    # Second convolutional layer -- maps 32 feature maps to 64.
    with tf.name_scope('conv2'):
        W_conv2 = weight_variable([5, 5, 32, 64])
        b_conv2 = bias_variable([64])
        h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)

    # Second pooling layer.
    with tf.name_scope('pool2'):
        h_pool2 = max_pool_2x2(h_conv2)

    # Fully connected layer 1 -- after 2 round of downsampling, our 28x28 image
    # is down to 7x7x64 feature maps -- maps this to 1024 features.
    with tf.name_scope('fc1'):
        W_fc1 = weight_variable([7 * 7 * 64, 1024])
        b_fc1 = bias_variable([1024])

        h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])
        h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

    # Dropout - controls the complexity of the model, prevents co-adaptation of
    # features.
    with tf.name_scope('dropout'):
        keep_prob = tf.placeholder(tf.float32)
        h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)

    # Map the 1024 features to 10 classes, one for each digit
    with tf.name_scope('fc2'):
        W_fc2 = weight_variable([1024, 10])
        b_fc2 = bias_variable([10])

        y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2
    return y_conv, keep_prob


def conv2d(x, W):
    """conv2d returns a 2d convolution layer with full stride."""
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')


def max_pool_2x2(x):
    """max_pool_2x2 downsamples a feature map by 2X."""
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
                          strides=[1, 2, 2, 1], padding='SAME')


def weight_variable(shape):
    """weight_variable generates a weight variable of a given shape."""
    initial = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(initial)


def bias_variable(shape):
    """bias_variable generates a bias variable of a given shape."""
    initial = tf.constant(0.1, shape=shape)
    return tf.Variable(initial)


def main(_):
    # Import data
    mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)

    # Create the model
    # x = tf.placeholder(tf.float32, [None, 784])   # 原始程序
    x = tf.placeholder(tf.float32, [None, 784], name="input")  # 修改后

    # Define loss and optimizer
    y_ = tf.placeholder(tf.float32, [None, 10])

    # Build the graph for the deep net
    y_conv, keep_prob = deepnn(x)

    with tf.name_scope('loss'):
        cross_entropy = tf.nn.softmax_cross_entropy_with_logits(labels=y_,
                                                                logits=y_conv)
    cross_entropy = tf.reduce_mean(cross_entropy)

    with tf.name_scope('adam_optimizer'):
        train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)

    with tf.name_scope('accuracy'):
        correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
        correct_prediction = tf.cast(correct_prediction, tf.float32)
    accuracy = tf.reduce_mean(correct_prediction)

    graph_location = tempfile.mkdtemp()
    print('Saving graph to: %s' % graph_location)
    train_writer = tf.summary.FileWriter(graph_location)
    train_writer.add_graph(tf.get_default_graph())

    saver = tf.train.Saver()  # 增加部分

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for i in range(2000):
            batch = mnist.train.next_batch(50)
            if i % 100 == 0:
                train_accuracy = accuracy.eval(feed_dict={
                    x: batch[0], y_: batch[1], keep_prob: 1.0})
                print('step %d, training accuracy %g' % (i, train_accuracy))
            train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5})

        print('test accuracy %g' % accuracy.eval(feed_dict={
            x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))

        graph_location = "."  # 增加部分
        save_path = saver.save(sess, graph_location + "/mnist_model")  # 增加部分


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--data_dir', type=str,
                        default='/tmp/tensorflow/mnist/input_data',
                        help='Directory for storing input data')
    FLAGS, unparsed = parser.parse_known_args()
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)

2、保存模型
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function

import argparse
import sys
import tempfile

import tensorflow as tf

FLAGS = None


def deepnn(x):
    with tf.name_scope('reshape'):
        x_image = tf.reshape(x, [-1, 28, 28, 1])

    with tf.name_scope('conv1'):
        W_conv1 = weight_variable([5, 5, 1, 32])
        b_conv1 = bias_variable([32])
        h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)

    with tf.name_scope('pool1'):
        h_pool1 = max_pool_2x2(h_conv1)

    with tf.name_scope('conv2'):
        W_conv2 = weight_variable([5, 5, 32, 64])
        b_conv2 = bias_variable([64])
        h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)

    with tf.name_scope('pool2'):
        h_pool2 = max_pool_2x2(h_conv2)

    with tf.name_scope('fc1'):
        W_fc1 = weight_variable([7 * 7 * 64, 1024])
        b_fc1 = bias_variable([1024])

        h_pool2_flat = tf.reshape(h_pool2, [-1, 7 * 7 * 64])
        h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)

    with tf.name_scope('fc2'):
        W_fc2 = weight_variable([1024, 10])
        b_fc2 = bias_variable([10])
        y_conv = tf.matmul(h_fc1, W_fc2) + b_fc2
    return y_conv


def conv2d(x, W):
    """conv2d returns a 2d convolution layer with full stride."""
    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')


def max_pool_2x2(x):
    """max_pool_2x2 downsamples a feature map by 2X."""
    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
                          strides=[1, 2, 2, 1], padding='SAME')


def weight_variable(shape):
    """weight_variable generates a weight variable of a given shape."""
    initial = tf.truncated_normal(shape, stddev=0.1)
    return tf.Variable(initial)


def bias_variable(shape):
    """bias_variable generates a bias variable of a given shape."""
    initial = tf.constant(0.1, shape=shape)
    return tf.Variable(initial)


def main(_):
    x = tf.placeholder(tf.float32, [None, 784], name="input")
    y_conv = deepnn(x)
    output = tf.nn.softmax(y_conv, name='output')

    saver = tf.train.Saver(tf.global_variables())

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        sess.run(tf.local_variables_initializer())
        saver.restore(sess, '.' + '/mnist_model')
        saver.save(sess, 'C:/Users/ym/Desktop/model' + '/mnist_inference')


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    parser.add_argument('--data_dir', type=str,
                        default='/tmp/tensorflow/mnist/input_data',
                        help='Directory for storing input data')
    FLAGS, unparsed = parser.parse_known_args()
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)

3、模型转换为中间件
模型存储后mnist_inference.index、mnist_inference.data - 00000 - of - 00001、mnist_inference.meta
将这三个文件通过TensorFlow自带的freezeing功能将模型存为.pb文件
将.pb文件转化为神经计算棒能够识别的模型
1、管理员身份运行cmd
2、cd
C:\Program
Files(x86)\IntelSWTools\openvino_2019
.2
.275\deployment_tools\model_optimizer
3、python
mo_tf.py - -input_model
C:\Users\ym\Desktop\model\model1.pb - -input_shape[1, 28, 28, 1] - -data_type
FP16
最后生成三个文件.bin文件.mapping文件.xml文件

4、C + +调用
# include 
# include 
# include 
# include 
# include "Windows.h"

# ifdef UNICODE
# include 
# endif

# include 
# include 
# include 

using
namespace
InferenceEngine;

# ifndef UNICODE
# define tcout std::cout
# define _T(STR) STR
# else
# define tcout std::wcout
# endif

# ifndef UNICODE
int
main()
{
    const
file_name_t
input_model
{"C:/Users/ym/Desktop/model/cnn_model.xml"};
const
file_name_t
input_image_path
{"C:/Users/ym/Desktop/22/q/Y_200402_26__17.jpg"};
const
std::string
device_name
{"MYRIAD"}; // 设备名字
MYRIAD是神经计算棒，CPU是电脑自带CPU

             // -----------------------------------------------------------------------------------------------------

             // --------------------------- 1.
Load
inference
engine
instance - ------------------------------------
Core
ie;
// -----------------------------------------------------------------------------------------------------

   // --------------------------- 2.
Read
IR
Generated
by
ModelOptimizer(.xml and.bin
files) ------------
CNNNetReader
network_reader;
network_reader.ReadNetwork(fileNameToString(input_model));
network_reader.ReadWeights(fileNameToString(input_model).substr(0, input_model.size() - 4) + ".bin");
network_reader.getNetwork().setBatchSize(1);
CNNNetwork
network = network_reader.getNetwork();

// -----------------------------------------------------------------------------------------------------

   // --------------------------- 3.
Configure
input & output - --------------------------------------------
// --------------------------- Prepare
input
blobs - ----------------------------------------------------
InputInfo::Ptr
input_info = network.getInputsInfo().begin()->second;
std::string
input_name = network.getInputsInfo().begin()->first;

/ *Mark
input as resizable
by
setting
of
a
resize
algorithm.
* In
this
case
we
will
be
able
to
set
an
input
blob
of
any
shape
to
an
infer
request.
* Resize and layout
conversions
are
executed
automatically
during
inference * /
input_info->getPreProcess().setResizeAlgorithm(RESIZE_BILINEAR);
input_info->setLayout(Layout::NHWC);
input_info->setPrecision(Precision::U8);

// --------------------------- Prepare
output
blobs - ---------------------------------------------------
DataPtr
output_info = network.getOutputsInfo().begin()->second;
std::string
output_name = network.getOutputsInfo().begin()->first;

output_info->setPrecision(Precision::FP32);
// -----------------------------------------------------------------------------------------------------

   // --------------------------- 4.
Loading
model
to
the
device - -----------------------------------------
ExecutableNetwork
executable_network = ie.LoadNetwork(network, device_name);
// -----------------------------------------------------------------------------------------------------

   // --------------------------- 5.
Create
infer
request - ------------------------------------------------
InferRequest
infer_request = executable_network.CreateInferRequest();
// -----------------------------------------------------------------------------------------------------

   // --------------------------- 6.
Prepare
input - -------------------------------------------------------
/ * Read
input
image
to
a
blob and set
it
to
an
infer
request
without
resize and layout
conversions. * /
cv::Mat
image = cv::imread(input_image_path, -1);

Blob::Ptr
imgBlob = wrapMat2Blob(image); // just
wrap
Mat
data
by
Blob::Ptr
without
allocating
of
new
memory
infer_request.SetBlob(input_name, imgBlob); // infer_request
accepts
input
blob
of
any
size

- -----------------------------------------------------------------

// --------------------------- 7.
Do
inference - -------------------------------------------------------
/ * Running
the
request
synchronously * /
infer_request.Infer();
// -----------------------------------------------------------------------------------------------------

   // --------------------------- 8.
Process
output - -----------------------------------------------------
Blob::Ptr
output2 = infer_request.GetBlob(output_name);
// Print
classification
results
ClassificationResult
classificationResult(output, {fileNameToString(input_image_path2)});
classificationResult.print();
// -----------------------------------------------------------------------------------------------------

return EXIT_SUCCESS;
}






参考
[1]
https: // docs.openvinotoolkit.org / latest / _docs_install_guides_installing_openvino_windows.html
[2]
https: // movidius.github.io / ncsdk / tf_compile_guidance.html