TensorFlow_Keras_ImageClassificatioin_01.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "TensorFlow_Keras_ImageClassificatioin_01.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.6"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/scpepper69/ml-learning-materials/blob/master/TensorFlow_Keras_ImageClassificatioin_01.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bYOnZhT30bGu",
        "colab_type": "text"
      },
      "source": [
        "# AI・機械学習 勉強会 #1\n",
        "## - 多層パーセプトロンおよび畳み込みニューラルネットワークによる画像分類 -\n",
        "\n",
        "\n",
        "※ まえがき\n",
        "\n",
        "このjupyter notebookは、東京大学 松尾研究所が公開している[DL4USコンテンツ](https://weblab.t.u-tokyo.ac.jp/dl4us/)を学習する中で、Deep Learningの前提知識が無い方や、数学が苦手な方においては、理論の理解が追い付かない部分があると感じ、トライ＆エラーによる実践を行うことがより理解を深めることができると考えて作成したものです。\n",
        "\n",
        "数式や理論を省略しているため、これらも含めて詳しく理解するには、[DL4USコンテンツ](https://weblab.t.u-tokyo.ac.jp/dl4us/)の1章および2章を合わせて参照すると良いかと思います。"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "nRUBscIZP3DA",
        "colab_type": "text"
      },
      "source": [
        "## 目次\n",
        "\n",
        "  \n",
        "1.1.   概要\n",
        "\n",
        "1.2.   実装プロセス\n",
        "\n",
        "1.   環境準備\n",
        "2.   データの準備\n",
        "3.   モデル構築\n",
        "4.   モデルの学習\n",
        "5.   モデルによる予測\n",
        "5.   中間層データの可視化\n",
        "5.   モデルのファイル出力\n",
        "\n",
        "\n",
        "\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aNaQV4x5qxii",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "##1.1 概要\n",
        "TensorFlow + Kerasによる実装を通じて基本的なDeep Learningの考え方や実装方法を学習します。\n",
        "\n",
        "また、画像分類における多層パーセプトロン(MLP)と畳込みニューラルネットワーク(CNN)での比較を行い、モデル構造の違いとチューニングパラメータについて理解を深めます。\n",
        "\n",
        "ソースコードはPython3を前提に書いてあります。また、特にCIFAR10においては学習に時間がかかるため、Google ColabではGPUランタイムを使ってください。\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J2n0HLjd0bG3",
        "colab_type": "text"
      },
      "source": [
        "\n",
        "\n",
        "---\n",
        "\n",
        "\n",
        "\n",
        "## 1.2 実装プロセス"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HVNVVU2o0bG5",
        "colab_type": "text"
      },
      "source": [
        "ここでは、画像分類をMLP(Multi Layer Perceptron)とCNN(Convolutional Neural Network)の2つモデルにいて、教師あり学習モデルの実装およびモデル構造、学習結果の比較をしてみます。\n",
        "\n",
        "実装に使用する題材は、MNIST(手書き数字画像)と、CIFAR10(乗り物や動物などの画像)の2種類とします。\n",
        "\n",
        "\n",
        "参考：https://keras.io/ja/datasets/"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kfinbX-8rfFr",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.1 環境準備\n",
        "\n",
        "プログラムを動かすにあたって、モデルおよびデータセットを指定してください。\n",
        "\n",
        "<モデル構造>\n",
        "* 多層パーセプトロン(MLP)\n",
        "* 畳込みニューラルネットワーク(CNN)\n",
        "\n",
        "<データセット>\n",
        "* MNIST：手書き数字(グレースケール)\n",
        "* CIFAR10：乗り物や動物の画像(カラー)\n",
        "\n",
        "以下の順序で動かしてみると良いかと思います。\n",
        "\n",
        "1. MNISTデータセットのMLP\n",
        "2. MNISTデータセットのCNN\n",
        "3. CIFAR10データセットのMLP\n",
        "4. CIFAR10データセットのCNN\n",
        "\n",
        "また、このjupyter notebookはTensorFlow 1.13.1を前提に実装しています。2019/6/14時点でバージョンが変更(1.14)となったため、ダウングレードしておきます。\n",
        "\n",
        "※その他ライブラリが変更になり、動作しなくなる可能性もありますが、Google Colabの仕様次第のため、都度チューニングして使ってください。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8b5TKT0SrZVa",
        "colab_type": "code",
        "outputId": "e9a94081-b202-46e4-d297-74c486d5882b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 649
        }
      },
      "source": [
        "# ライブラリのバージョン調整\n",
        "# 2019/6/14時点でデフォルトのTensorflowバージョンが変わった模様\n",
        "!pip install tensorflow-estimator==1.13.0 tensorflow-gpu==1.13.1\n",
        "\n",
        "# モデル構造を指定します (MLP or CNN)\n",
        "model_structure=\"CNN\"\n",
        "# CNNの場に指定 (CNN, VGG16, RESNET1 or RESNET2)\n",
        "model_opt=\"RESNET2\"\n",
        "# データセットを指定します (MNIST or CIFAR10)\n",
        "data_set=\"CIFAR10\""
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting tensorflow-estimator==1.13.0\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/bb/48/13f49fc3fa0fdf916aa1419013bb8f2ad09674c275b4046d5ee669a46873/tensorflow_estimator-1.13.0-py2.py3-none-any.whl (367kB)\n",
            "\u001b[K     |████████████████████████████████| 368kB 6.8MB/s \n",
            "\u001b[?25hCollecting tensorflow-gpu==1.13.1\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/7b/b1/0ad4ae02e17ddd62109cd54c291e311c4b5fd09b4d0678d3d6ce4159b0f0/tensorflow_gpu-1.13.1-cp36-cp36m-manylinux1_x86_64.whl (345.2MB)\n",
            "\u001b[K     |████████████████████████████████| 345.2MB 64kB/s \n",
            "\u001b[?25hRequirement already satisfied: absl-py>=0.1.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-estimator==1.13.0) (0.7.1)\n",
            "Collecting mock>=2.0.0 (from tensorflow-estimator==1.13.0)\n",
            "  Downloading https://files.pythonhosted.org/packages/05/d2/f94e68be6b17f46d2c353564da56e6fb89ef09faeeff3313a046cb810ca9/mock-3.0.5-py2.py3-none-any.whl\n",
            "Requirement already satisfied: numpy>=1.13.3 in /usr/local/lib/python3.6/dist-packages (from tensorflow-estimator==1.13.0) (1.16.4)\n",
            "Requirement already satisfied: six>=1.10.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-estimator==1.13.0) (1.12.0)\n",
            "Requirement already satisfied: keras-applications>=1.0.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (1.0.8)\n",
            "Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (0.33.4)\n",
            "Requirement already satisfied: grpcio>=1.8.6 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (1.15.0)\n",
            "Collecting tensorboard<1.14.0,>=1.13.0 (from tensorflow-gpu==1.13.1)\n",
            "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/0f/39/bdd75b08a6fba41f098b6cb091b9e8c7a80e1b4d679a581a0ccd17b10373/tensorboard-1.13.1-py3-none-any.whl (3.2MB)\n",
            "\u001b[K     |████████████████████████████████| 3.2MB 41.4MB/s \n",
            "\u001b[?25hRequirement already satisfied: gast>=0.2.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (0.2.2)\n",
            "Requirement already satisfied: keras-preprocessing>=1.0.5 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (1.1.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (1.1.0)\n",
            "Requirement already satisfied: protobuf>=3.6.1 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (3.7.1)\n",
            "Requirement already satisfied: astor>=0.6.0 in /usr/local/lib/python3.6/dist-packages (from tensorflow-gpu==1.13.1) (0.8.0)\n",
            "Requirement already satisfied: h5py in /usr/local/lib/python3.6/dist-packages (from keras-applications>=1.0.6->tensorflow-gpu==1.13.1) (2.8.0)\n",
            "Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.6/dist-packages (from tensorboard<1.14.0,>=1.13.0->tensorflow-gpu==1.13.1) (0.15.4)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.6/dist-packages (from tensorboard<1.14.0,>=1.13.0->tensorflow-gpu==1.13.1) (3.1.1)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.6/dist-packages (from protobuf>=3.6.1->tensorflow-gpu==1.13.1) (41.0.1)\n",
            "\u001b[31mERROR: tensorflow 1.14.0 has requirement tensorboard<1.15.0,>=1.14.0, but you'll have tensorboard 1.13.1 which is incompatible.\u001b[0m\n",
            "\u001b[31mERROR: tensorflow 1.14.0 has requirement tensorflow-estimator<1.15.0rc0,>=1.14.0rc0, but you'll have tensorflow-estimator 1.13.0 which is incompatible.\u001b[0m\n",
            "Installing collected packages: mock, tensorflow-estimator, tensorboard, tensorflow-gpu\n",
            "  Found existing installation: tensorflow-estimator 1.14.0\n",
            "    Uninstalling tensorflow-estimator-1.14.0:\n",
            "      Successfully uninstalled tensorflow-estimator-1.14.0\n",
            "  Found existing installation: tensorboard 1.14.0\n",
            "    Uninstalling tensorboard-1.14.0:\n",
            "      Successfully uninstalled tensorboard-1.14.0\n",
            "Successfully installed mock-3.0.5 tensorboard-1.13.1 tensorflow-estimator-1.13.0 tensorflow-gpu-1.13.1\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UZaJgfxvyLAy",
        "colab_type": "text"
      },
      "source": [
        "Google Colabratoryは、ランタイムが初期化されるとデータも失われます。\n",
        "\n",
        "学習した中のチェックポイントが学習済みモデルを再利用できるよう、Google Driveをマウントし、ここに出力できるようしておきましょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bf6gYTI1xt0n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os, shutil\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive/')\n",
        "gdrive_base='/content/drive/My Drive/Colab Notebooks/'\n",
        "\n",
        "# TensorBorad用ログ\n",
        "log_dir=gdrive_base+'ImageClassification/logs/'\n",
        "if not os.path.exists(log_dir):\n",
        "    os.makedirs(log_dir)\n",
        "\n",
        "# チェックポイントおよび学習済みモデルファイル\n",
        "model_dir=gdrive_base+'ImageClassification/model/'\n",
        "if not os.path.exists(model_dir):\n",
        "    os.makedirs(model_dir)\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "CFKVp80t0bG8",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.2 データの準備\n",
        "教師あり学習を行う場合、学習させるデータを用意する必要があります。\n",
        "\n",
        "MNISTおよびCIFAR10はKerasにて提供されていますので、ここではこちらを使います。\n",
        "\n",
        "なお、データの収集は非常に地味かつ大変な作業ですが、このデータの事前準備が非常に重要であり、\"使える\"データの量や種類が、モデルの精度に大きく関わってきます。\n",
        "\n",
        "今回は用意されたものを使いますが、自分で集めたデータセットにて学習させることも非常に良い経験となるため、ぜひご自身でもトライしてみてください。\n",
        "         \n",
        "今回扱うMNISTおよびCIFAR10のデータ構造は下記となっています。\n",
        "\n",
        "* x：画像(MNIST：28×28x1、CIFAR10：32x32x1) ※縦x横xグレー(1)orカラー(3)\n",
        "* y：正解のラベル(xの画像が表す意味)\n",
        "\n",
        "また、これらデータセットは学習用と評価用に最初から分割されています。\n",
        "\n",
        "* (x_train, y_train):モデルの学習用\n",
        "* (x_test, y_test):モデルの評価用\n",
        "\n",
        "Deep Learningでは学習データでの認識精度が高くても意味がありません。\n",
        "\n",
        "学習時にはなかった未知のデータについても正しく認識させること(これを汎化性能といいます)が目的であり、これが精度の高い学習モデルと言えます。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MIjKuxuD0bG-",
        "colab_type": "code",
        "outputId": "214add71-7ae3-4d49-c30c-c6904db896dc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "# Kerasにて用意されているデータセットをダウンロードします\n",
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.datasets import cifar10\n",
        "\n",
        "# MNIST\n",
        "if data_set == \"MNIST\":\n",
        "    (x_train, y_train), (x_test, y_test) = mnist.load_data()\n",
        "    img_size,img_color=28,1\n",
        "# CIFAR10\n",
        "else:\n",
        "    (x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "    img_size,img_color=32,3"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz\n",
            "170500096/170498071 [==============================] - 6s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M3pSTYqA0bHI",
        "colab_type": "text"
      },
      "source": [
        "ダウンロードした画像と画像に対応したラベルを表示してみます。\n",
        "\n",
        "画像表示には、matplotlibライブラリを用います。これはPythonにてグラフ表示によく使われるライブラリなので、使用方法は覚えておくと良いです。\n",
        "\n",
        "参考：https://matplotlib.org/api/pyplot_api.html\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T-XaTkOc0bHL",
        "colab_type": "code",
        "outputId": "28d9fe56-6d87-4373-95eb-3648e49a80b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 478
        }
      },
      "source": [
        "#jupyter notebook用マジックコマンド\n",
        "%matplotlib inline \n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "fig = plt.figure(figsize=(9, 15))\n",
        "fig.subplots_adjust(left=0, right=1, bottom=0, top=0.5, hspace=0.05, wspace=0.05)\n",
        "\n",
        "# 各MNIST画像の上に（タイトルとして）対応するラベルを表示\n",
        "for i in range(30):\n",
        "    ax = fig.add_subplot(3, 10, i + 1, xticks=[], yticks=[])\n",
        "    ax.set_title(str(y_train[i]))\n",
        "    ax.imshow(x_train[i], cmap='gray')"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAp0AAAHNCAYAAABcs8BUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
            "text/plain": [
              "<Figure size 648x1080 with 30 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_ZCxZxgl0bHX",
        "colab_type": "text"
      },
      "source": [
        "これから実施するのは、MINISTにしても、CIFAR10にしても、いわゆる10クラス分類の問題です。（分類すべき対象のことを**クラス**と呼びます）\n",
        "\n",
        "画像分類を行うにあたって以下2点、前提として理解しておく必要があります。\n",
        "\n",
        "1.   ラベルの意味\n",
        "\n",
        "        MNISTにおいては、ラベルの値が数字としての意味と一致しているように見えますが、これはただの偶然でしかなく、あくまで、ラベルの値はクラスのインデックスに過ぎない数値ということです。\n",
        "\n",
        "        CIFAR10で試してみるとわかりますが、飛行機の画像のラベルは\"0\"であってairplaneという意味を持っているわけではありません。\n",
        "\n",
        "        要するに、Deep Learningとして理解している(させている)わけではないということです。\n",
        "\n",
        "        後ほど実装内にも出てきますが、各インデックスの意味づけは人がプログラム内でつけることなります。\n",
        "\n",
        "        画像分類の問題に限らず、現在のDeep Learningにおいては、1次的な結果を学習させ人間を超える精度を発揮することは可能ですが、2次的情報を付加するのはあくまで人間であり、その情報を理解し拡張することはまだできないと認識しておいてください。\n",
        "\n",
        "\n",
        "2.   ラベルの表現\n",
        "\n",
        "        もう一つ、インデックスなので、データとしては数字で表現されてしまいますが、Deep Learningを行う上で、これを数字としての意味を持って扱ってもらっては困るため、全クラス中のどのクラスに該当するかという表現に置き換える必要があります。\n",
        "\n",
        "        これを**one-hot表現**と呼んでいます。\n",
        "\n",
        "        * ラベル 0 = airplane -> (one-hot表現) -> [1,0,0,0,0,0,0,0,0,0]  ※10要素の配列の0番目に該当\n",
        "        * ラベル 1 = car -> (one-hot表現) -> [0,1,0,0,0,0,0,0,0,0] ※10要素の配列の1番目に該当\n",
        "\n",
        "    Kerasには、この**one-hot表現**に変換してくれる関数(keras.utils.to_categorical)が用意されています。\n",
        "    "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DBvnPZCU0bHZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.utils import to_categorical\n",
        "import numpy as np\n",
        "\n",
        "# ラベルデータをone-hot表現へ変換 (10クラス)\n",
        "y_train = to_categorical(y_train, 10)\n",
        "y_test = to_categorical(y_test, 10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7WDd1fFQ0eYP",
        "colab_type": "text"
      },
      "source": [
        "MLPでは、画像上の位置情報を意識せず、１次元配列として次の層に全結合させていきます。\n",
        "\n",
        "一方、CNNは画像の位置情報を保持した３次元配列のまま、ConvolutionおよびPoolingすることで特徴量を抽出します。\n",
        "\n",
        "そのため、学習および推論する際のINPUTデータ形式は、MLPは１次元配列、CNNは３次元配列となります。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AfE9xtc-0Mgo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# 画像データの形式をMLPおよびCNNに適した形に修正\n",
        "# for CNN \n",
        "if model_structure==\"CNN\":\n",
        "    x_train = x_train.reshape(-1, img_size, img_size, img_color).astype(np.float32)\n",
        "    x_test = x_test.reshape(-1, img_size, img_size, img_color).astype(np.float32)\n",
        "    input_shape = (img_size, img_size, img_color)\n",
        "# for MLP        \n",
        "else:\n",
        "    # 入力画像を行列(28x28)からベクトル(長さ784)に変換\n",
        "    x_train = x_train.reshape(-1, img_size*img_size*img_color)\n",
        "    x_test = x_test.reshape(-1, img_size*img_size*img_color)\n",
        "    input_shape = (img_size*img_size*img_color,)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bUUkN6qj0bHi",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.3 モデル構築\n",
        "\n",
        "これから、学習モデルを構築します。\n",
        "\n",
        "モデルの構築は、モデルの「容器」としてSquential()を実施したのち、その中に順にレイヤーを追加していく流れになります。\n",
        "\n",
        "* [Sequential](https://keras.io/ja/models/sequential/)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1rIGYTTK0bHk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# モデル構築用ライブラリをインポート\n",
        "from tensorflow import keras\n",
        "from tensorflow.keras.models import Sequential, Model\n",
        "from tensorflow.keras.layers import Dense, Activation, Flatten\n",
        "\n",
        "# CNNモデル用ライブラリ\n",
        "from tensorflow.python.keras.layers import Conv2D, Convolution2D, MaxPooling2D, Dropout, BatchNormalization, GlobalAveragePooling2D,AveragePooling2D,Input\n",
        "from tensorflow.python.keras import initializers\n",
        "from tensorflow.keras.applications.vgg16 import VGG16\n",
        "from tensorflow.keras.regularizers import l2\n",
        "\n",
        "# モデルの「容器」を作成\n",
        "model = Sequential()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vWgvceiHbKnb",
        "colab_type": "text"
      },
      "source": [
        "これから、モデル構造を定義していきます。ここでは、モデルの切り替えが容易にできるように、ここではモデル構造ごとにfunctionを用意しておきます。\n",
        "\n",
        "なお、Kerasでは最初にaddするレイヤにてInput_shapeを定義してやる必要があります。\n",
        "\n",
        "下記実装においては、入出力を除いた中間層について、MLPは全結合2層、CNNは畳込み2層＋全結合1層としています。\n",
        "\n",
        "各レイヤーのパラメータ等についてはKerasのドキュメントを参照してください。\n",
        "\n",
        "* [Dense/Actication/Dropout](https://keras.io/ja/layers/core/)\n",
        "* [Conv2D](https://keras.io/ja/layers/convolutional/)\n",
        "* [MaxPooling](https://keras.io/ja/layers/pooling/)\n",
        "* [Batch Normalization](https://keras.io/ja/layers/normalization/)\n",
        "\n",
        "\n",
        "MLPにおいては、3つの異なる活性化関数(Activation)のfunctionを用意しています。\n",
        "\n",
        "また、CNNにおいては、Dropoutの有無とBatch Nrmalizationを実装したfunctionを用意しています。\n",
        "\n",
        "特にCIFAR10の分類においては、このままの実装では高い精度は得られないはずです。\n",
        "\n",
        "それぞれのモデルの比較とともに、層の深さやレイヤーの構成、レイヤーごとのパラメータなどを自由にいじってみて、どのような変化が起こるか試してみましょう。\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i8TOq83u0bHz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# MLPモデル\n",
        "def mlp_relu_model():\n",
        "    model.add(Dense(units=256, input_shape=input_shape))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Dense(units=100))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n",
        "\n",
        "def mlp_sigmoid_model():\n",
        "    model.add(Dense(units=256, input_shape=input_shape))\n",
        "    model.add(Activation('sigmoid'))\n",
        "    model.add(Dense(units=100))\n",
        "    model.add(Activation('sigmoid'))\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n",
        "\n",
        "def mlp_tanh_model():\n",
        "    model.add(Dense(units=256, input_shape=input_shape))\n",
        "    model.add(Activation('tanh'))\n",
        "    model.add(Dense(units=100))\n",
        "    model.add(Activation('tanh'))\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FLLXeRM3f7Fi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# CNNモデル\n",
        "def cnn_model():\n",
        "    model.add(Conv2D(32, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros(), input_shape=input_shape))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Conv2D(64, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(256, kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n",
        "\n",
        "# CNNモデル w/Dropout\n",
        "def cnn_w_dropout():\n",
        "    model.add(Conv2D(32, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros(), input_shape=input_shape))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Conv2D(64, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "    model.add(Dropout(0.5))\n",
        "\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(256, kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Dropout(0.5))\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n",
        "    \n",
        "# CNNモデル w/Batch Normalization\n",
        "def cnn_w_batchnorm():\n",
        "    model.add(Conv2D(32, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros(), input_shape=input_shape))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Conv2D(32, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "    model.add(Conv2D(64, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros(), input_shape=input_shape))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(Conv2D(64, (3, 3), kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(256, kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    model.add(Activation('relu'))\n",
        "    model.add(BatchNormalization())\n",
        "    model.add(Dense(units=10))\n",
        "    model.add(Activation('softmax'))\n",
        "\n",
        "    return model\n",
        "    "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nt0Xe1505Quo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# VGG16モデル\n",
        "def cnn_vgg16():\n",
        "    input_tensor = Input(shape=(img_size, img_size, img_color))\n",
        "    vgg16 = VGG16(include_top=False, weights='imagenet', input_tensor=input_tensor)    \n",
        "    \n",
        "    add_model = Sequential()\n",
        "    add_model.add(Flatten())\n",
        "    add_model.add(Dense(256, kernel_initializer=initializers.TruncatedNormal(stddev=0.1), bias_initializer=initializers.Zeros()))\n",
        "    add_model.add(Activation('relu'))\n",
        "    add_model.add(BatchNormalization())\n",
        "    add_model.add(Dense(units=10))\n",
        "    add_model.add(Activation('softmax'))\n",
        "    \n",
        "    vgg16_model = Model(inputs=vgg16.input, outputs=add_model(vgg16.output))\n",
        "\n",
        "    #fix weights VGG16 layers\n",
        "    for layer in vgg16.layers:\n",
        "        layer.trainable = False\n",
        "        \n",
        "    return vgg16_model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OGWk7WWk5Ya3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# ResNetモデル from Keras Documentation\n",
        "\n",
        "# Model parameter\n",
        "# ----------------------------------------------------------------------------\n",
        "#           |      | 200-epoch | Orig Paper| 200-epoch | Orig Paper| sec/epoch\n",
        "# Model     |  n   | ResNet v1 | ResNet v1 | ResNet v2 | ResNet v2 | GTX1080Ti\n",
        "#           |v1(v2)| %Accuracy | %Accuracy | %Accuracy | %Accuracy | v1 (v2)\n",
        "# ----------------------------------------------------------------------------\n",
        "# ResNet20  | 3 (2)| 92.16     | 91.25     | -----     | -----     | 35 (---)\n",
        "# ResNet32  | 5(NA)| 92.46     | 92.49     | NA        | NA        | 50 ( NA)\n",
        "# ResNet44  | 7(NA)| 92.50     | 92.83     | NA        | NA        | 70 ( NA)\n",
        "# ResNet56  | 9 (6)| 92.71     | 93.03     | 93.01     | NA        | 90 (100)\n",
        "# ResNet110 |18(12)| 92.65     | 93.39+-.16| 93.15     | 93.63     | 165(180)\n",
        "# ResNet164 |27(18)| -----     | 94.07     | -----     | 94.54     | ---(---)\n",
        "# ResNet1001| (111)| -----     | 92.39     | -----     | 95.08+-.14| ---(---)\n",
        "# ---------------------------------------------------------------------------\n",
        "n = 3\n",
        "\n",
        "# Model version\n",
        "# Orig paper: version = 1 (ResNet v1), Improved ResNet: version = 2 (ResNet v2)\n",
        "version = 1\n",
        "\n",
        "# Computed depth from supplied model parameter n\n",
        "#if version == 1:\n",
        "#    depth = n * 6 + 2\n",
        "#elif version == 2:\n",
        "#    depth = n * 9 + 2\n",
        "\n",
        "# ResNetモデル\n",
        "def resnet_layer(inputs,\n",
        "                 num_filters=16,\n",
        "                 kernel_size=3,\n",
        "                 strides=1,\n",
        "                 activation='relu',\n",
        "                 batch_normalization=True,\n",
        "                 conv_first=True):\n",
        "    \"\"\"2D Convolution-Batch Normalization-Activation stack builder\n",
        "\n",
        "    # Arguments\n",
        "        inputs (tensor): input tensor from input image or previous layer\n",
        "        num_filters (int): Conv2D number of filters\n",
        "        kernel_size (int): Conv2D square kernel dimensions\n",
        "        strides (int): Conv2D square stride dimensions\n",
        "        activation (string): activation name\n",
        "        batch_normalization (bool): whether to include batch normalization\n",
        "        conv_first (bool): conv-bn-activation (True) or\n",
        "            bn-activation-conv (False)\n",
        "\n",
        "    # Returns\n",
        "        x (tensor): tensor as input to the next layer\n",
        "    \"\"\"\n",
        "    conv = Conv2D(num_filters,\n",
        "                  kernel_size=kernel_size,\n",
        "                  strides=strides,\n",
        "                  padding='same',\n",
        "                  kernel_initializer='he_normal',\n",
        "                  kernel_regularizer=l2(1e-4))\n",
        "\n",
        "    x = inputs\n",
        "    if conv_first:\n",
        "        x = conv(x)\n",
        "        if batch_normalization:\n",
        "            x = BatchNormalization()(x)\n",
        "        if activation is not None:\n",
        "            x = Activation(activation)(x)\n",
        "    else:\n",
        "        if batch_normalization:\n",
        "            x = BatchNormalization()(x)\n",
        "        if activation is not None:\n",
        "            x = Activation(activation)(x)\n",
        "        x = conv(x)\n",
        "    return x\n",
        "\n",
        "\n",
        "def resnet_v1(input_shape, depth=20, num_classes=10):\n",
        "    \"\"\"ResNet Version 1 Model builder [a]\n",
        "\n",
        "    Stacks of 2 x (3 x 3) Conv2D-BN-ReLU\n",
        "    Last ReLU is after the shortcut connection.\n",
        "    At the beginning of each stage, the feature map size is halved (downsampled)\n",
        "    by a convolutional layer with strides=2, while the number of filters is\n",
        "    doubled. Within each stage, the layers have the same number filters and the\n",
        "    same number of filters.\n",
        "    Features maps sizes:\n",
        "    stage 0: 32x32, 16\n",
        "    stage 1: 16x16, 32\n",
        "    stage 2:  8x8,  64\n",
        "    The Number of parameters is approx the same as Table 6 of [a]:\n",
        "    ResNet20 0.27M\n",
        "    ResNet32 0.46M\n",
        "    ResNet44 0.66M\n",
        "    ResNet56 0.85M\n",
        "    ResNet110 1.7M\n",
        "\n",
        "    # Arguments\n",
        "        input_shape (tensor): shape of input image tensor\n",
        "        depth (int): number of core convolutional layers\n",
        "        num_classes (int): number of classes (CIFAR10 has 10)\n",
        "\n",
        "    # Returns\n",
        "        model (Model): Keras model instance\n",
        "    \"\"\"\n",
        "    if (depth - 2) % 6 != 0:\n",
        "        raise ValueError('depth should be 6n+2 (eg 20, 32, 44 in [a])')\n",
        "    # Start model definition.\n",
        "    num_filters = 16\n",
        "    num_res_blocks = int((depth - 2) / 6)\n",
        "\n",
        "    inputs = Input(shape=input_shape)\n",
        "    x = resnet_layer(inputs=inputs)\n",
        "    # Instantiate the stack of residual units\n",
        "    for stack in range(3):\n",
        "        for res_block in range(num_res_blocks):\n",
        "            strides = 1\n",
        "            if stack > 0 and res_block == 0:  # first layer but not first stack\n",
        "                strides = 2  # downsample\n",
        "            y = resnet_layer(inputs=x,\n",
        "                             num_filters=num_filters,\n",
        "                             strides=strides)\n",
        "            y = resnet_layer(inputs=y,\n",
        "                             num_filters=num_filters,\n",
        "                             activation=None)\n",
        "            if stack > 0 and res_block == 0:  # first layer but not first stack\n",
        "                # linear projection residual shortcut connection to match\n",
        "                # changed dims\n",
        "                x = resnet_layer(inputs=x,\n",
        "                                 num_filters=num_filters,\n",
        "                                 kernel_size=1,\n",
        "                                 strides=strides,\n",
        "                                 activation=None,\n",
        "                                 batch_normalization=False)\n",
        "            x = keras.layers.add([x, y])\n",
        "            x = Activation('relu')(x)\n",
        "        num_filters *= 2\n",
        "\n",
        "    # Add classifier on top.\n",
        "    # v1 does not use BN after last shortcut connection-ReLU\n",
        "    x = AveragePooling2D(pool_size=8)(x)\n",
        "    y = Flatten()(x)\n",
        "    outputs = Dense(num_classes,\n",
        "                    activation='softmax',\n",
        "                    kernel_initializer='he_normal')(y)\n",
        "\n",
        "    # Instantiate model.\n",
        "    model = Model(inputs=inputs, outputs=outputs)\n",
        "    return model\n",
        "\n",
        "\n",
        "def resnet_v2(input_shape, depth=29, num_classes=10):\n",
        "    \"\"\"ResNet Version 2 Model builder [b]\n",
        "\n",
        "    Stacks of (1 x 1)-(3 x 3)-(1 x 1) BN-ReLU-Conv2D or also known as\n",
        "    bottleneck layer\n",
        "    First shortcut connection per layer is 1 x 1 Conv2D.\n",
        "    Second and onwards shortcut connection is identity.\n",
        "    At the beginning of each stage, the feature map size is halved (downsampled)\n",
        "    by a convolutional layer with strides=2, while the number of filter maps is\n",
        "    doubled. Within each stage, the layers have the same number filters and the\n",
        "    same filter map sizes.\n",
        "    Features maps sizes:\n",
        "    conv1  : 32x32,  16\n",
        "    stage 0: 32x32,  64\n",
        "    stage 1: 16x16, 128\n",
        "    stage 2:  8x8,  256\n",
        "\n",
        "    # Arguments\n",
        "        input_shape (tensor): shape of input image tensor\n",
        "        depth (int): number of core convolutional layers\n",
        "        num_classes (int): number of classes (CIFAR10 has 10)\n",
        "\n",
        "    # Returns\n",
        "        model (Model): Keras model instance\n",
        "    \"\"\"\n",
        "    if (depth - 2) % 9 != 0:\n",
        "        raise ValueError('depth should be 9n+2 (eg 56 or 110 in [b])')\n",
        "    # Start model definition.\n",
        "    num_filters_in = 16\n",
        "    num_res_blocks = int((depth - 2) / 9)\n",
        "\n",
        "    inputs = Input(shape=input_shape)\n",
        "    # v2 performs Conv2D with BN-ReLU on input before splitting into 2 paths\n",
        "    x = resnet_layer(inputs=inputs,\n",
        "                     num_filters=num_filters_in,\n",
        "                     conv_first=True)\n",
        "\n",
        "    # Instantiate the stack of residual units\n",
        "    for stage in range(3):\n",
        "        for res_block in range(num_res_blocks):\n",
        "            activation = 'relu'\n",
        "            batch_normalization = True\n",
        "            strides = 1\n",
        "            if stage == 0:\n",
        "                num_filters_out = num_filters_in * 4\n",
        "                if res_block == 0:  # first layer and first stage\n",
        "                    activation = None\n",
        "                    batch_normalization = False\n",
        "            else:\n",
        "                num_filters_out = num_filters_in * 2\n",
        "                if res_block == 0:  # first layer but not first stage\n",
        "                    strides = 2    # downsample\n",
        "\n",
        "            # bottleneck residual unit\n",
        "            y = resnet_layer(inputs=x,\n",
        "                             num_filters=num_filters_in,\n",
        "                             kernel_size=1,\n",
        "                             strides=strides,\n",
        "                             activation=activation,\n",
        "                             batch_normalization=batch_normalization,\n",
        "                             conv_first=False)\n",
        "            y = resnet_layer(inputs=y,\n",
        "                             num_filters=num_filters_in,\n",
        "                             conv_first=False)\n",
        "            y = resnet_layer(inputs=y,\n",
        "                             num_filters=num_filters_out,\n",
        "                             kernel_size=1,\n",
        "                             conv_first=False)\n",
        "            if res_block == 0:\n",
        "                # linear projection residual shortcut connection to match\n",
        "                # changed dims\n",
        "                x = resnet_layer(inputs=x,\n",
        "                                 num_filters=num_filters_out,\n",
        "                                 kernel_size=1,\n",
        "                                 strides=strides,\n",
        "                                 activation=None,\n",
        "                                 batch_normalization=False)\n",
        "            x = keras.layers.add([x, y])\n",
        "\n",
        "        num_filters_in = num_filters_out\n",
        "\n",
        "    # Add classifier on top.\n",
        "    # v2 has BN-ReLU before Pooling\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Activation('relu')(x)\n",
        "    x = AveragePooling2D(pool_size=8)(x)\n",
        "    y = Flatten()(x)\n",
        "    outputs = Dense(num_classes,\n",
        "                    activation='softmax',\n",
        "                    kernel_initializer='he_normal')(y)\n",
        "\n",
        "    # Instantiate model.\n",
        "    model = Model(inputs=inputs, outputs=outputs)\n",
        "    return model\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3DNVsEupdMjn",
        "colab_type": "text"
      },
      "source": [
        "それでは、任意のモデルを選び、構築してみましょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Nq7BxPyz0bH3",
        "colab_type": "code",
        "outputId": "c03b8639-14b8-4186-ce3f-fd7564affb99",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        }
      },
      "source": [
        "# model構築\n",
        "if model_structure==\"CNN\":\n",
        "    if model_opt==\"VGG16\":\n",
        "        model = cnn_vgg16()\n",
        "    elif model_opt==\"RESNET1\":\n",
        "        model = resnet_v1(input_shape=input_shape)\n",
        "    elif model_opt==\"RESNET2\":\n",
        "        model = resnet_v2(input_shape=input_shape)\n",
        "    else:\n",
        "        model=cnn_model()\n",
        "#        model=cnn_w_dropout()\n",
        "#        model=cnn_w_batchnorm()\n",
        "else:\n",
        "    mlp_relu_model()\n",
        "#    model=mlp_sigmoid_model()\n",
        "#    model=mlp_tanh_model()\n"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/resource_variable_ops.py:435: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sSEVpmwMo1U8",
        "colab_type": "text"
      },
      "source": [
        "先ほど定義したモデルを視覚的に理解できるよう、モデル構造を図示してみます。\n",
        "\n",
        "図示するツールはconvert-drawerというツールを用います。\n",
        "\n",
        "convert-drawerのオリジナルのgithubはココですが、MLPやDropoutやBatch Normalizationへも対応できるよう、修正を加えています。\n",
        "\n",
        "また、ついでにに各レイヤーの定義情報を参考として出力してみます。この定義情報は各レイヤーのadd時にパラメータとして付与できるものたちです。\n",
        "\n",
        "デフォルトとして明示していないものがどれだけあり、実際はどのようなパラメータで動いているか見てみてください。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vBmICHhQOX57",
        "colab_type": "code",
        "outputId": "32d9d6b9-3b53-46f0-b7aa-1b011473ec37",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 153
        }
      },
      "source": [
        "if os.path.exists('./convert_drawer'):\n",
        "    shutil.rmtree('./convert_drawer')\n",
        "\n",
        "!git clone https://github.com/scpepper69/convnet-drawer.git convert_drawer\n",
        "import sys\n",
        "sys.path.append('./convert_drawer/')\n",
        "from keras_util import convert_drawer_model\n",
        "from keras_models import AlexNet\n",
        "from matplotlib_util import save_model_to_file\n",
        "from IPython.display import *\n",
        "\n",
        "# RESNETのスキップ構造はツール未対応のため\n",
        "if not \"RESNET\" in model_opt:\n",
        "    # get Keras sequential model\n",
        "    #keras_sequential_model = AlexNet.get_model()\n",
        "    draw_model = convert_drawer_model(model)\n",
        "\n",
        "    # save as svg file\n",
        "    draw_model.save_fig(f'{model_dir}{model_structure}_{data_set}_model_structure.svg')\n",
        "    display_svg(SVG(f'{model_dir}{model_structure}_{data_set}_model_structure.svg'))"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Cloning into 'convert_drawer'...\n",
            "remote: Enumerating objects: 102, done.\u001b[K\n",
            "remote: Counting objects: 100% (102/102), done.\u001b[K\n",
            "remote: Compressing objects: 100% (71/71), done.\u001b[K\n",
            "remote: Total 321 (delta 62), reused 66 (delta 31), pack-reused 219\u001b[K\n",
            "Receiving objects: 100% (321/321), 1.58 MiB | 3.29 MiB/s, done.\n",
            "Resolving deltas: 100% (192/192), done.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E5Dz7x4Scrf8",
        "colab_type": "text"
      },
      "source": [
        "学習の開始にあたり、最後にcompile関数でコンパイルを行います。\n",
        "\n",
        "compile関数でも、学習にあたって以下のパラメータを指定する必要があります。\n",
        "\n",
        "* optimizer（最適化手法）\n",
        "* loss（損失関数）\n",
        "* metrics（評価関数（任意））\n",
        "\n",
        "以下サンプルコードでは、現在Kerasで利用できるoptimizerを並べてみました。こちらもそれぞれどのような結果になるか比較してみるといいでしょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OI4UDcU60bH7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras import optimizers\n",
        "from tensorflow.keras import losses\n",
        "# モデルの学習方法について指定しておく\n",
        "model.compile(loss='categorical_crossentropy', optimizer=optimizers.SGD(lr=0.001, momentum=0.9), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.Adam(lr=0.001), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.RMSprop(lr=0.001), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.Adagrad(lr=0.001), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.Adadelta(lr=0.001), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.Adamax(lr=0.001), metrics=['accuracy'])\n",
        "#model.compile(loss='categorical_crossentropy', optimizer=optimizers.Nadam(lr=0.001), metrics=['accuracy'])\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1ZwzQE-FYxdR",
        "colab_type": "text"
      },
      "source": [
        "Optimizerアルゴリズムの動作イメージ(作者：Alec Radford)\n",
        "\n",
        "\n",
        "<img src=\"http://sebastianruder.com/content/images/2016/01/contours_evaluation_optimizers.gif\" width=\"400\">\n",
        "<img src=\"http://sebastianruder.com/content/images/2016/01/saddle_point_evaluation_optimizers.gif\" width=\"400\">\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5LYqpAQPO04I",
        "colab_type": "text"
      },
      "source": [
        "早速学習開始と行きたいところですが、学習の途中結果を記録する(Checkpoint)ことと、TensorBoard向けのログ情報を取得できるようにしておきましょう。\n",
        "\n",
        "モデル構造や結果等の視覚化することで、何がどのように動いているのか、理解の手助けとなります。\n",
        "\n",
        "このjupyter notebook上でも可視化ロジックを組み込んでいますが、通常のモデル開発においては、TensorBoardを用いるほうがスマートでしょう。\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "weO_9YIVJ4dA",
        "colab_type": "text"
      },
      "source": [
        "TensorBoardでも参照できるよう、設定を組み込んでおきます。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s80FRVRL0bH-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# TensorBoardでの可視化のため、出力先の設定\n",
        "from tensorflow.keras import callbacks\n",
        "\n",
        "tb_cb = callbacks.TensorBoard(log_dir=log_dir, histogram_freq=1,write_images=1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GJVjFnA8h5LT",
        "colab_type": "text"
      },
      "source": [
        "チェックポイントを生成しておくことで、中断した学習の再開や、チェックポイントを利用した静的モデルの出力を行うことができます。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Rk3ohoKJhd9P",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# チェックポイント出力先\n",
        "RUN = RUN + 1 if 'RUN' in locals() else 1\n",
        "checkpoint_path = model_dir + f'run{RUN}/' + model_structure + \"_\" + data_set + \"_cp-{epoch:04d}.ckpt\"\n",
        "checkpoint_dir = os.path.dirname(checkpoint_path)\n",
        "\n",
        "# チェックポイントコールバックを作る\n",
        "cp_cb = callbacks.ModelCheckpoint(checkpoint_path, save_weights_only=True, verbose=1, period=5)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DMWlqNgvPBvp",
        "colab_type": "text"
      },
      "source": [
        "学習の実行の前に、モデルのサマリ情報を確認してみましょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "135NNq6iO-b2",
        "colab_type": "code",
        "outputId": "3054a251-b39e-43dd-9813-5a28b6aa8332",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# モデルのサマリ情報の表示\n",
        "model.summary()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            (None, 32, 32, 3)    0                                            \n",
            "__________________________________________________________________________________________________\n",
            "conv2d (Conv2D)                 (None, 32, 32, 16)   448         input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1 (BatchNo (None, 32, 32, 16)   64          conv2d[0][0]                     \n",
            "__________________________________________________________________________________________________\n",
            "activation (Activation)         (None, 32, 32, 16)   0           batch_normalization_v1[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_1 (Conv2D)               (None, 32, 32, 16)   272         activation[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_1 (Batch (None, 32, 32, 16)   64          conv2d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_1 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_1[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_2 (Conv2D)               (None, 32, 32, 16)   2320        activation_1[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_2 (Batch (None, 32, 32, 16)   64          conv2d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_2 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_2[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_4 (Conv2D)               (None, 32, 32, 64)   1088        activation[0][0]                 \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_3 (Conv2D)               (None, 32, 32, 64)   1088        activation_2[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "add (Add)                       (None, 32, 32, 64)   0           conv2d_4[0][0]                   \n",
            "                                                                 conv2d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_3 (Batch (None, 32, 32, 64)   256         add[0][0]                        \n",
            "__________________________________________________________________________________________________\n",
            "activation_3 (Activation)       (None, 32, 32, 64)   0           batch_normalization_v1_3[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_5 (Conv2D)               (None, 32, 32, 16)   1040        activation_3[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_4 (Batch (None, 32, 32, 16)   64          conv2d_5[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_4 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_4[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_6 (Conv2D)               (None, 32, 32, 16)   2320        activation_4[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_5 (Batch (None, 32, 32, 16)   64          conv2d_6[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_5 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_5[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_7 (Conv2D)               (None, 32, 32, 64)   1088        activation_5[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "add_1 (Add)                     (None, 32, 32, 64)   0           add[0][0]                        \n",
            "                                                                 conv2d_7[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_6 (Batch (None, 32, 32, 64)   256         add_1[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_6 (Activation)       (None, 32, 32, 64)   0           batch_normalization_v1_6[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_8 (Conv2D)               (None, 32, 32, 16)   1040        activation_6[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_7 (Batch (None, 32, 32, 16)   64          conv2d_8[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_7 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_7[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_9 (Conv2D)               (None, 32, 32, 16)   2320        activation_7[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_8 (Batch (None, 32, 32, 16)   64          conv2d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "activation_8 (Activation)       (None, 32, 32, 16)   0           batch_normalization_v1_8[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_10 (Conv2D)              (None, 32, 32, 64)   1088        activation_8[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "add_2 (Add)                     (None, 32, 32, 64)   0           add_1[0][0]                      \n",
            "                                                                 conv2d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_9 (Batch (None, 32, 32, 64)   256         add_2[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_9 (Activation)       (None, 32, 32, 64)   0           batch_normalization_v1_9[0][0]   \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_11 (Conv2D)              (None, 16, 16, 64)   4160        activation_9[0][0]               \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_10 (Batc (None, 16, 16, 64)   256         conv2d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_10 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_10[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_12 (Conv2D)              (None, 16, 16, 64)   36928       activation_10[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_11 (Batc (None, 16, 16, 64)   256         conv2d_12[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_11 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_11[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_14 (Conv2D)              (None, 16, 16, 128)  8320        add_2[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_13 (Conv2D)              (None, 16, 16, 128)  8320        activation_11[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_3 (Add)                     (None, 16, 16, 128)  0           conv2d_14[0][0]                  \n",
            "                                                                 conv2d_13[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_12 (Batc (None, 16, 16, 128)  512         add_3[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_12 (Activation)      (None, 16, 16, 128)  0           batch_normalization_v1_12[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_15 (Conv2D)              (None, 16, 16, 64)   8256        activation_12[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_13 (Batc (None, 16, 16, 64)   256         conv2d_15[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_13 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_13[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_16 (Conv2D)              (None, 16, 16, 64)   36928       activation_13[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_14 (Batc (None, 16, 16, 64)   256         conv2d_16[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_14 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_14[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_17 (Conv2D)              (None, 16, 16, 128)  8320        activation_14[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_4 (Add)                     (None, 16, 16, 128)  0           add_3[0][0]                      \n",
            "                                                                 conv2d_17[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_15 (Batc (None, 16, 16, 128)  512         add_4[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_15 (Activation)      (None, 16, 16, 128)  0           batch_normalization_v1_15[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_18 (Conv2D)              (None, 16, 16, 64)   8256        activation_15[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_16 (Batc (None, 16, 16, 64)   256         conv2d_18[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_16 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_16[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_19 (Conv2D)              (None, 16, 16, 64)   36928       activation_16[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_17 (Batc (None, 16, 16, 64)   256         conv2d_19[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_17 (Activation)      (None, 16, 16, 64)   0           batch_normalization_v1_17[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_20 (Conv2D)              (None, 16, 16, 128)  8320        activation_17[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_5 (Add)                     (None, 16, 16, 128)  0           add_4[0][0]                      \n",
            "                                                                 conv2d_20[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_18 (Batc (None, 16, 16, 128)  512         add_5[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_18 (Activation)      (None, 16, 16, 128)  0           batch_normalization_v1_18[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_21 (Conv2D)              (None, 8, 8, 128)    16512       activation_18[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_19 (Batc (None, 8, 8, 128)    512         conv2d_21[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_19 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_19[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_22 (Conv2D)              (None, 8, 8, 128)    147584      activation_19[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_20 (Batc (None, 8, 8, 128)    512         conv2d_22[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_20 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_20[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_24 (Conv2D)              (None, 8, 8, 256)    33024       add_5[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_23 (Conv2D)              (None, 8, 8, 256)    33024       activation_20[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_6 (Add)                     (None, 8, 8, 256)    0           conv2d_24[0][0]                  \n",
            "                                                                 conv2d_23[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_21 (Batc (None, 8, 8, 256)    1024        add_6[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_21 (Activation)      (None, 8, 8, 256)    0           batch_normalization_v1_21[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_25 (Conv2D)              (None, 8, 8, 128)    32896       activation_21[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_22 (Batc (None, 8, 8, 128)    512         conv2d_25[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_22 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_22[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_26 (Conv2D)              (None, 8, 8, 128)    147584      activation_22[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_23 (Batc (None, 8, 8, 128)    512         conv2d_26[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_23 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_23[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_27 (Conv2D)              (None, 8, 8, 256)    33024       activation_23[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_7 (Add)                     (None, 8, 8, 256)    0           add_6[0][0]                      \n",
            "                                                                 conv2d_27[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_24 (Batc (None, 8, 8, 256)    1024        add_7[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_24 (Activation)      (None, 8, 8, 256)    0           batch_normalization_v1_24[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_28 (Conv2D)              (None, 8, 8, 128)    32896       activation_24[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_25 (Batc (None, 8, 8, 128)    512         conv2d_28[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_25 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_25[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_29 (Conv2D)              (None, 8, 8, 128)    147584      activation_25[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_26 (Batc (None, 8, 8, 128)    512         conv2d_29[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "activation_26 (Activation)      (None, 8, 8, 128)    0           batch_normalization_v1_26[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "conv2d_30 (Conv2D)              (None, 8, 8, 256)    33024       activation_26[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "add_8 (Add)                     (None, 8, 8, 256)    0           add_7[0][0]                      \n",
            "                                                                 conv2d_30[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_v1_27 (Batc (None, 8, 8, 256)    1024        add_8[0][0]                      \n",
            "__________________________________________________________________________________________________\n",
            "activation_27 (Activation)      (None, 8, 8, 256)    0           batch_normalization_v1_27[0][0]  \n",
            "__________________________________________________________________________________________________\n",
            "average_pooling2d (AveragePooli (None, 1, 1, 256)    0           activation_27[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "flatten (Flatten)               (None, 256)          0           average_pooling2d[0][0]          \n",
            "__________________________________________________________________________________________________\n",
            "dense (Dense)                   (None, 10)           2570        flatten[0][0]                    \n",
            "==================================================================================================\n",
            "Total params: 849,002\n",
            "Trainable params: 843,786\n",
            "Non-trainable params: 5,216\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sUmiS87vdonR",
        "colab_type": "text"
      },
      "source": [
        "学習途上のチェックポイントから再開させることもできます。\n",
        "\n",
        "load_weights関数にてチェックポイントファイルから重みをロードします。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dAQ_PWeidls-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# チェックポイントから学習済みパラメータを復元\n",
        "#model.load_weights(f'{model_dir}run1/{model_structure}_{data_set}_cp-0010.ckpt')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6_yBnQiN0bID",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.4 モデルの学習\n",
        "\n",
        "ようやく学習に入ります。学習処理はfit関数もしくはfit_generatorの2種類ありますが、ここではfit関数を使います。\n",
        "\n",
        "ここでは以下の引数を与えます。\n",
        "\n",
        "* x_train：学習に使用する画像データ\n",
        "* y_train：学習に使用するラベルデータ(one-hot表現)\n",
        "* batch_size：学習中のパラメータ更新を1回行うにあたって用いるサンプル数（ミニバッチのサイズ）\n",
        "* epochs：学習のエポック数\n",
        "* verbose：学習のログを出力するか（0:しない、1：バーで出力、2:エポックごとに出力）\n",
        "* validation_data：検証用データ\n",
        "* callbacks：訓練中のモデルの挙動を監視できるcallback関数を指定できます\n",
        "\n",
        "fit関数にはこれ以外にもパラメータを持っています。\n",
        "\n",
        "* [fit](https://keras.io/ja/models/model/#fit)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "scrolled": false,
        "id": "_IQUEb790bIE",
        "colab_type": "code",
        "outputId": "687bf63f-5f35-4f92-e252-95c22b8f7638",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 666
        }
      },
      "source": [
        "# epoch数を指定\n",
        "epochs=10\n",
        "\n",
        "# batchサイズを指定\n",
        "batch_size=500\n",
        "\n",
        "# 学習の実行\n",
        "result = model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, verbose=1,callbacks=[tb_cb, cp_cb], validation_data=(x_test, y_test))"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "Epoch 1/10\n",
            "50000/50000 [==============================] - 844s 17ms/sample - loss: 2.8728 - acc: 0.1875 - val_loss: 3.1290 - val_acc: 0.1399\n",
            "Epoch 2/10\n",
            "50000/50000 [==============================] - 829s 17ms/sample - loss: 2.7295 - acc: 0.2527 - val_loss: 2.7112 - val_acc: 0.2488\n",
            "Epoch 3/10\n",
            "50000/50000 [==============================] - 829s 17ms/sample - loss: 2.6465 - acc: 0.2815 - val_loss: 2.6399 - val_acc: 0.2901\n",
            "Epoch 4/10\n",
            "50000/50000 [==============================] - 828s 17ms/sample - loss: 2.5821 - acc: 0.2996 - val_loss: 2.5725 - val_acc: 0.3086\n",
            "Epoch 5/10\n",
            "49500/50000 [============================>.] - ETA: 0s - loss: 2.5312 - acc: 0.3130\n",
            "Epoch 00005: saving model to /content/drive/My Drive/Colab Notebooks/ImageClassification/model/run1/CNN_CIFAR10_cp-0005.ckpt\n",
            "WARNING:tensorflow:This model was compiled with a Keras optimizer (<tensorflow.python.keras.optimizers.SGD object at 0x7f0613df32e8>) but is being saved in TensorFlow format with `save_weights`. The model's weights will be saved, but unlike with TensorFlow optimizers in the TensorFlow format the optimizer's state will not be saved.\n",
            "\n",
            "Consider using a TensorFlow optimizer from `tf.train`.\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/network.py:1436: update_checkpoint_state (from tensorflow.python.training.checkpoint_management) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.train.CheckpointManager to manage checkpoints rather than manually editing the Checkpoint proto.\n",
            "50000/50000 [==============================] - 815s 16ms/sample - loss: 2.5314 - acc: 0.3132 - val_loss: 2.5147 - val_acc: 0.3256\n",
            "Epoch 6/10\n",
            "50000/50000 [==============================] - 814s 16ms/sample - loss: 2.4913 - acc: 0.3240 - val_loss: 2.4759 - val_acc: 0.3355\n",
            "Epoch 7/10\n",
            "50000/50000 [==============================] - 818s 16ms/sample - loss: 2.4585 - acc: 0.3349 - val_loss: 2.4377 - val_acc: 0.3456\n",
            "Epoch 8/10\n",
            "50000/50000 [==============================] - 825s 17ms/sample - loss: 2.4309 - acc: 0.3443 - val_loss: 2.4119 - val_acc: 0.3514\n",
            "Epoch 9/10\n",
            "50000/50000 [==============================] - 823s 16ms/sample - loss: 2.4071 - acc: 0.3507 - val_loss: 2.3898 - val_acc: 0.3584\n",
            "Epoch 10/10\n",
            "49500/50000 [============================>.] - ETA: 0s - loss: 2.3855 - acc: 0.3591\n",
            "Epoch 00010: saving model to /content/drive/My Drive/Colab Notebooks/ImageClassification/model/run1/CNN_CIFAR10_cp-0010.ckpt\n",
            "WARNING:tensorflow:This model was compiled with a Keras optimizer (<tensorflow.python.keras.optimizers.SGD object at 0x7f0613df32e8>) but is being saved in TensorFlow format with `save_weights`. The model's weights will be saved, but unlike with TensorFlow optimizers in the TensorFlow format the optimizer's state will not be saved.\n",
            "\n",
            "Consider using a TensorFlow optimizer from `tf.train`.\n",
            "50000/50000 [==============================] - 816s 16ms/sample - loss: 2.3853 - acc: 0.3589 - val_loss: 2.3714 - val_acc: 0.3632\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XCLIozIF0bIJ",
        "colab_type": "text"
      },
      "source": [
        "学習結果の評価については、evaluate関数にて得ることができます。\n",
        "\n",
        "主な引数は次の通りです。\n",
        "\n",
        "* x：評価に使用する入力データ\n",
        "* y：評価に使用する出力データ\n",
        "* batch_size：1回の評価を行うにあたって用いるサンプル数\n",
        "* verbose：評価のログを出力するか（0:しない、1：する(デフォルト)）\n",
        "\n",
        "基本的には、損失(Loss)は低ければ低いほうが、評価(Accuracy)は高ければ高いほうが良いです。\n",
        "\n",
        "Accuracyはモデルの精度そのもの、Lossは学習が効率よく行われているかを示す指標で、Lossが高いまま収束していかない＝効率が良くなく、モデル構造やパラメータ改善の余地あり、という感覚でよいかと思います。\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ys0GtCR60bIK",
        "colab_type": "code",
        "outputId": "0e003ae1-9f19-48c5-a9a8-03d54d2af77f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "score = model.evaluate(x_test, y_test, verbose=0)\n",
        "print('Test loss:', score[0])\n",
        "print('Test accuracy:', score[1])"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Test loss: 2.3714390647888184\n",
            "Test accuracy: 0.3632\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uvQIQuhhWdOo",
        "colab_type": "text"
      },
      "source": [
        "epochごとのAccuracyおよびLossの遷移をグラフ化してみましょう。\n",
        "\n",
        "fit関数のreturnから、データを取得することができます。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UB8isRg_hfGT",
        "colab_type": "code",
        "outputId": "375ad47f-2312-4ec7-b8ed-f0ac1292667c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 590
        }
      },
      "source": [
        "result.history.keys() # ヒストリデータのラベルを見てみる\n",
        "plt.plot(range(1, epochs+1), result.history['acc'], label=\"training\")\n",
        "plt.plot(range(1, epochs+1), result.history['val_acc'], label=\"validation\")\n",
        "plt.title('Accuracy History')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Accuracy')\n",
        "plt.legend()\n",
        "plt.xlim([1,epochs])\n",
        "plt.ylim([0,1])\n",
        "plt.show()\n",
        "plt.savefig(model_structure+'_'+data_set+'_'+'acc.png')\n",
        "\n",
        "plt.plot(range(1, epochs+1), result.history['loss'], label=\"training\")\n",
        "plt.plot(range(1, epochs+1), result.history['val_loss'], label=\"validation\")\n",
        "plt.title('Loss History')\n",
        "plt.xlabel('Epochs')\n",
        "plt.ylabel('Loss')\n",
        "plt.legend()\n",
        "plt.xlim([1,epochs])\n",
        "plt.ylim([0,20])\n",
        "plt.show()\n",
        "plt.savefig(model_structure+'_'+data_set+'_'+'loss.png')"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nh5mZZeLgMDOzTP4/N4R6eiR2bjYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Zq7fmArk0bIN",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.5 モデルによる予測\n",
        "\n",
        "精度だけ見ても面白くないので、検証用データの予測結果を見てみましょう。\n",
        "\n",
        "Sequential.predict関数によって予測を行うことができます。\n",
        "\n",
        "主な引数は次の通りです。\n",
        "\n",
        "* x_test：予測に使用する入力データ\n",
        "* batch_size：まとめて1度に予測を行うサンプル数\n",
        "* verbose：評価のログを出力するか（0:しない(デフォルト)、1：する）"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "m7CJSTL50bIN",
        "colab_type": "code",
        "outputId": "ff4e842a-cd23-487b-9d68-198b3c9e65a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "classes = model.predict(x_test, batch_size=128, verbose=1)"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10000/10000 [==============================] - 4s 413us/sample\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kx85EWpfW7Nz",
        "colab_type": "text"
      },
      "source": [
        "上記コマンドで検証用データ10000件の予測を実施しています。\n",
        "\n",
        "どのような結果が得られているかイメージしやすくするために、画像を1枚抽出して、その結果を実際に見てみます。\n",
        "\n",
        "なお、モデル構築の際に、いずれのモデルにおいても、出力層をDense(10)、Activation('softmax')としていました。\n",
        "\n",
        "これは、10クラス(Dense(10))に、それぞれの確からしさを出力(Activation('softmax'))する、という意味になります。\n",
        "\n",
        "softmaxでは、すべてのクラスの出力値の合計が1になります。そのため、この値が確からしさそのものを指し、最も値の大きいラベルが推論結果として採用することとなります。\n",
        "\n",
        "なお、前述(1.2.2)にてラベルの意味づけに言及していましたが、本サンプルでは、ここでラベルの数値をラベルの意味に変換させています。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5sTY0plvOe-2",
        "colab_type": "code",
        "outputId": "c54dca94-cb46-41a6-9065-b0c7e991bd3f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        }
      },
      "source": [
        "# データセットの推論結果と元画像を確認\n",
        "# test_numを0～9999で指定してください\n",
        "%matplotlib inline\n",
        "test_num=1112\n",
        "test_img = np.squeeze(x_test[test_num])\n",
        "if not model_structure==\"CNN\":\n",
        "    test_img = np.squeeze(test_img.reshape(img_size, img_size, img_color).astype(np.float32))\n",
        "\n",
        "if data_set==\"MNIST\":\n",
        "    labels = np.array([0,1,2,3,4,5,6,7,8,9])\n",
        "else:\n",
        "    labels = np.array([\n",
        "        'airplane',\n",
        "        'automobile',\n",
        "        'bird',\n",
        "        'cat',\n",
        "        'deer',\n",
        "        'dog',\n",
        "        'frog',\n",
        "        'horse',\n",
        "        'ship',\n",
        "        'truck'])\n",
        "\n",
        "print(\"クラスごとの確からしさ(%)：\"+str(np.round(classes[test_num],decimals=2)*100))\n",
        "\n",
        "print(\"推論結果：\"+str(labels[classes[test_num].argmax()]))\n",
        "plt.imshow(test_img.astype(np.int),'gray')\n",
        "plt.show()"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "クラスごとの確からしさ(%)：[ 2.        2.       15.000001 21.       18.       17.       13.\n",
            "  7.        1.        2.      ]\n",
            "推論結果：cat\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIucXIlH+P0CYWKCeA0z4AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_B2-3A2nVfuj",
        "colab_type": "text"
      },
      "source": [
        "正しく予測されていたでしょうか？\n",
        "\n",
        "さらに進んで、検証用データにも存在しない画像で推論させてみましょう。\n",
        "\n",
        "手書き画像(MNISTの場合)や、Web検索で画像を取得し、Google Colabにアップロードしてみましょう。\n",
        "\n",
        "私が用意したサンプルデータも用意しておきます。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oZ8gMryWuxfh",
        "colab_type": "code",
        "outputId": "5e5f87bb-c946-4b5f-ef12-1740795a0e07",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 425
        }
      },
      "source": [
        "# サンプルデータのダウンロード\n",
        "!wget https://raw.githubusercontent.com/scpepper69/ml-learning-materials/master/sample/images001.zip\n",
        "!unzip images001.zip -d sample"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-07-05 08:26:33--  https://raw.githubusercontent.com/scpepper69/ml-learning-materials/master/sample/images001.zip\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 1230178 (1.2M) [application/zip]\n",
            "Saving to: ‘images001.zip’\n",
            "\n",
            "\rimages001.zip         0%[                    ]       0  --.-KB/s               \rimages001.zip       100%[===================>]   1.17M  --.-KB/s    in 0.03s   \n",
            "\n",
            "2019-07-05 08:26:34 (36.9 MB/s) - ‘images001.zip’ saved [1230178/1230178]\n",
            "\n",
            "Archive:  images001.zip\n",
            "  inflating: sample/airplane001.png  \n",
            "  inflating: sample/car001.png       \n",
            "  inflating: sample/eight001.png     \n",
            "  inflating: sample/five001.png      \n",
            "  inflating: sample/four001.png      \n",
            "  inflating: sample/nine001.png      \n",
            "  inflating: sample/one001.png       \n",
            "  inflating: sample/seven001.png     \n",
            "  inflating: sample/six001.png       \n",
            "  inflating: sample/three001.png     \n",
            "  inflating: sample/two001.png       \n",
            "  inflating: sample/zero001.png      \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "E7ITflj5bE2t",
        "colab_type": "text"
      },
      "source": [
        "ファイルをアップロードし終えたら、ファイル名を指定して推論してみましょう。\n",
        "\n",
        "img_path変数にアップロードしたファイル名を指定し、実行してみてください。\n",
        "\n",
        "こちらも正しい結果が得られたでしょうか。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uDEx3_iRNpGK",
        "colab_type": "code",
        "outputId": "7d8a35fb-630b-4305-eaa2-3846618340ab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 337
        }
      },
      "source": [
        "import cv2\n",
        "    \n",
        "# Google Colabにファイルをアップロード\n",
        "# アップロードしたファイルを指定\n",
        "img_path=\"./sample/car001.png\"\n",
        "\n",
        "if data_set==\"MNIST\":\n",
        "    img = cv2.imread(img_path,cv2.IMREAD_GRAYSCALE)\n",
        "    img = 255 - img #白黒反転\n",
        "    plt.imshow(img, 'gray')\n",
        "else:\n",
        "    img = cv2.imread(img_path,cv2.IMREAD_COLOR)\n",
        "    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)    \n",
        "    plt.imshow(img)\n",
        "\n",
        "    \n",
        "img = cv2.resize(img, dsize=(img_size,img_size)) # 28 x 28にサイズ変更\n",
        "img = img.astype(np.float32)\n",
        "print(img.size)\n",
        "\n",
        "if model_structure==\"CNN\":\n",
        "    img = img.reshape(1, img_size, img_size ,img_color) # 2次元から4次元に変換\n",
        "else:    \n",
        "    img = img.reshape(1, img_size*img_size*img_color).astype(np.float32)\n",
        "\n",
        "pred = model.predict(img,verbose=1)\n",
        "\n",
        "# 推論処理\n",
        "print(\"クラスごとの確からしさ(%)：\"+str(np.round(pred,decimals=2)*100))\n",
        "print(\"推論結果：\"+str(labels[pred.argmax()]))"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "3072\n",
            "\r1/1 [==============================] - 0s 54ms/sample\n",
            "クラスごとの確からしさ(%)：[[ 1. 62.  1.  2.  1.  3.  3.  7.  1. 20.]]\n",
            "推論結果：automobile\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQUAAAD8CAYAAAB+fLH0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pQ4aUVfFVYfE",
        "colab_type": "text"
      },
      "source": [
        "予測精度の最終確認として、Confusion Matrixを表示してみます。\n",
        "\n",
        "X軸とY軸が一致する箇所に集中していれば(要するに左上から右下に斜めに赤くなっている)、精度の良いモデルと言えるでしょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BIOawSjjUl0y",
        "colab_type": "code",
        "outputId": "2188e003-0613-4262-865a-e958c4a9047c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 478
        }
      },
      "source": [
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "from sklearn.metrics import confusion_matrix\n",
        "from sklearn.metrics import classification_report\n",
        "\n",
        "#print(classification_report(np.argmax(y_test, 1), np.argmax(classes, 1)))\n",
        "#print(confusion_matrix(np.argmax(y_test, 1), np.argmax(classes, 1)))\n",
        "\n",
        "cmatrix = confusion_matrix(np.argmax(y_test, 1), np.argmax(classes, 1))\n",
        "cmatrix_plt = pd.DataFrame(cmatrix, index=labels, columns=labels)\n",
        "\n",
        "plt.figure(figsize = (10,7))\n",
        "sns.heatmap(cmatrix_plt, annot=True, cmap=\"Reds\", fmt=\"d\")\n",
        "plt.show()"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAlQAAAHNCAYAAADYGeU/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x504 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "D8ZcKoTgbbCf",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.6 中間層データの可視化\n",
        "\n",
        "ここまでで、データを準備し、モデルを構築し、学習、推論させることができました。\n",
        "\n",
        "どのように実装し、どのように使うのか、簡単にイメージすることができたのではないでしょうか。\n",
        "\n",
        "この項では、モデルがどのように動きているか理解を深めるために、中間層のデータを見てみます。\n",
        "\n",
        "ただ、MLPでは1次元に変換してしまっているので、見てもよくわからないかもしれませんが。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lVxresNopyLM",
        "colab_type": "code",
        "outputId": "1b37b9f2-4f94-42ae-9aed-0cb2e999094f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        }
      },
      "source": [
        "from tensorflow.keras import backend as K\n",
        "import matplotlib.pyplot as plt\n",
        "from PIL import Image\n",
        "\n",
        "# 中間層の特徴マップを出力する\n",
        "def middle_layer_output(numlay, INPDATA):\n",
        "    get_layer = K.function([model.layers[0].input], [model.layers[numlay].output])\n",
        "    # InputLayerは除外(VGG16転移学習)\n",
        "    # print(str(model.layers[numlay].__class__.__name__))\n",
        "    if str(model.layers[numlay].__class__.__name__) == \"InputLayer\":\n",
        "        return\n",
        "    layer_name = model.layers[numlay].name\n",
        "    layer_output_comp = get_layer([INPDATA])[0]\n",
        "    if layer_output_comp.ndim == 4:\n",
        "        dim_n = layer_output_comp.shape[0]\n",
        "        dim_x = layer_output_comp.shape[1]\n",
        "        dim_y = layer_output_comp.shape[2]\n",
        "        dim_z = layer_output_comp.shape[3]\n",
        "    else:\n",
        "        dim_n = layer_output_comp.shape[0]\n",
        "        dim_z = layer_output_comp.shape[1]\n",
        "\n",
        "    #中間層の出力を図化する\n",
        "    plt.figure(figsize=(16,512))\n",
        "    for num_pic in range(dim_n):\n",
        "        if layer_output_comp.ndim == 4:\n",
        "            for num_map in range(dim_z):\n",
        "                img = layer_output_comp[num_pic, :, :, num_map]\n",
        "                img = Image.fromarray(np.uint8(img))\n",
        "                plt.subplot(128,8,num_map+1)\n",
        "                plt.axis('off')\n",
        "                plt.title(str(layer_name+':'+str(num_map+1)))\n",
        "                plt.imshow(img, cmap='gray')\n",
        "#                plt.imshow(img)\n",
        "    \n",
        "        else:\n",
        "            dims = int(len(layer_output_comp[0]))\n",
        "            img = layer_output_comp[num_pic, :]\n",
        "            img = img.reshape(1,dims).astype(np.float32)\n",
        "            plt.axis('off')\n",
        "            plt.title(str(layer_name))\n",
        "            plt.imshow(img, cmap='gray')\n",
        "#            plt.imshow(img)\n",
        "\n",
        "# レイヤーごとの特徴マップの出力\n",
        "for layers in range(len(model.layers)):\n",
        "    middle_layer_output(layers, img)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/matplotlib/pyplot.py:514: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).\n",
            "  max_open_warning, RuntimeWarning)\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q26KDv1Icumj",
        "colab_type": "text"
      },
      "source": [
        "CNNでは、画像の形式で扱っているので、ConvolutionおよびMaxPoolingにおいてはINPUT画像の変化が見て取れます。\n",
        "\n",
        "直感的に理解できるものではありませんが、特徴量をどのように捉え推論しているか、理解の一助にはなるかと思います。\n",
        "\n",
        "なお、Deep Learningモデルにおいては、INPUTに対する最終的なOUTPUTだけでなく、中間層のOUTPUTを自由に取得することができるという点も非常に重要な特性であることは覚えておいてください。\n",
        "\n",
        "また、OUTPUTの観点だけでなく、INPUTも同様で、中間層にINPUTし、OUTPUTを得ることもできます。\n",
        "\n",
        "シーケンシャルなモデル構造でなく、複雑なモデルを構築していくにあたって重要な特性です。\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iIXwHVGJ0bIP",
        "colab_type": "text"
      },
      "source": [
        "参考までに、jupyter notebook上でモデル構造など可視化させてきましたが、TensorBoardでもどのように見えるのか試してみましょう。\n",
        "\n",
        "Google Colabでは、直接HTTPアクセスはできないので、ngrokを用いて参照させます。\n",
        "\n",
        "下記コード実行後し、出力されるURLにアクセスしてみましょう。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Pyx_anHq0bIQ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Web参照のため、ngrokを利用\n",
        "if not os.path.exists('./ngrok'):\n",
        "    !wget https://bin.equinox.io/c/4VmDzA7iaHb/ngrok-stable-linux-amd64.zip\n",
        "    !unzip ngrok-stable-linux-amd64.zip\n",
        "\n",
        "# TensorBoardおよびngrokの起動\n",
        "import subprocess\n",
        "cmd = f'tensorboard --logdir=\"{log_dir}\" --host 0.0.0.0 --port 6006 &'\n",
        "proc_tb = subprocess.call(cmd, shell=True)\n",
        "\n",
        "cmd = \"./ngrok http 6006 &\"\n",
        "proc_ng = subprocess.call(cmd, shell=True)\n",
        "\n",
        "# TensorBoard URL\n",
        "!curl -s http://localhost:4040/api/tunnels | python -c \"import sys, json; print(json.load(sys.stdin)['tunnels'][0]['public_url'])\"\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nH37JEOU0bIU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# TensorBoardプロセスの停止\n",
        "#!ps -ef | grep tensorboard | grep -v grep | awk '{print \"kill -9\",$2}'| sh\n",
        "\n",
        "# ngrokプロセスの停止\n",
        "#!ps -ef | grep ngrok | grep -v grep | awk '{print \"kill -9\",$2}'| sh\n",
        "!mkdir model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sqbPWg1mfbA3",
        "colab_type": "text"
      },
      "source": [
        "### 1.2.7 モデルのファイル出力\n",
        "\n",
        "学習させたモデルを出力し、静的学習済みモデルとして外部で活用することもできます。\n",
        "\n",
        "ここでは、Keras形式に加え、活用の汎用性を考えてTensorFlowのプロトコルバッファ形式にて出力してみます。\n",
        "\n",
        "静的学習済みモデルは、プログラムで読み込んで使うことも、GraphPipeやTensorFlow Serving等のモデルサーバーを用いてAPI化して利用することもできます。\n",
        "\n",
        "Google ColabではDockerを扱うことができないので、ローカルPCやクラウド環境等で試してみてください。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_tcWm1vmlVvA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.python.framework import graph_util\n",
        "from tensorflow.python.framework import graph_io\n",
        "from tensorflow import identity\n",
        "\n",
        "# 学習セッションからグラフを復元\n",
        "ksess = K.get_session()\n",
        "graph = ksess.graph\n",
        "kgraph = graph.as_graph_def()\n",
        "\n",
        "# Outputノード名を取得\n",
        "num_output = 1\n",
        "prefix = \"output\"\n",
        "pred = [None]*num_output\n",
        "outputName = [None]*num_output\n",
        "for i in range(num_output):\n",
        "    outputName[i] = prefix + str(i)\n",
        "    pred[i] = identity(model.get_output_at(i), name=outputName[i])\n",
        "print('output name: ', outputName)\n",
        "\n",
        "# 変数を定数に変換\n",
        "constant_graph = graph_util.convert_variables_to_constants(ksess, ksess.graph.as_graph_def(), outputName)\n",
        "\n",
        "output_dir = model_dir\n",
        "output_graph_name = f\"{model_structure}_{model_opt}_{data_set}_{epochs}_frozen_graph.pb\"\n",
        "\n",
        "# プロトコルバッファ形式で静的学習済みモデルを出力\n",
        "graph_io.write_graph(constant_graph, output_dir, output_graph_name, as_text=False)\n",
        "\n",
        "# Keras形式でモデルを出力\n",
        "output_keras_name = f\"{model_dir}{model_structure}_{model_opt}_{data_set}_{epochs}_frozen_graph.h5\"\n",
        "model.save(output_keras_name, include_optimizer=False)\n"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}