7.Cifar10 feature map.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "kernelspec": {
      "display_name": "Python [conda env:tensorflow_p36]",
      "language": "python",
      "name": "conda-env-tensorflow_p36-py"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.4"
    },
    "colab": {
      "name": "7.+Cifar10_feature_map.ipynb",
      "provenance": []
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "Lz5ddkInbF80",
        "colab_type": "code",
        "colab": {},
        "outputId": "55b857d7-18b9-4c5b-edb1-1aa90d5d1d58"
      },
      "source": [
        "from __future__ import print_function\n",
        "import keras\n",
        "from keras.datasets import cifar10\n",
        "from keras.preprocessing.image import ImageDataGenerator\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Dropout, Activation, Flatten\n",
        "from keras.layers.normalization import BatchNormalization\n",
        "from keras.layers import Conv2D, MaxPooling2D\n",
        "from keras.regularizers import l2\n",
        "import numpy as np\n",
        "import os\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/site-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
            "  from ._conv import register_converters as _register_converters\n",
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7ZWX_ZBObF9H",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Defining the parameters\n",
        "batch_size = 32\n",
        "num_classes = 10\n",
        "epochs = 50"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "63vID6jJbF9T",
        "colab_type": "code",
        "colab": {},
        "outputId": "d36b64e3-605a-4989-e7d7-d21d5dab3a6f"
      },
      "source": [
        "# Splitting the data between train and test\n",
        "(x_train, y_train), (x_test, y_test) = cifar10.load_data()\n",
        "print('x_train shape:', x_train.shape)\n",
        "print(x_train.shape[0], 'train samples')\n",
        "print(x_test.shape[0], 'test samples')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "x_train shape: (50000, 32, 32, 3)\n",
            "50000 train samples\n",
            "10000 test samples\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UxDGQyHRbF9i",
        "colab_type": "code",
        "colab": {},
        "outputId": "84248a5b-717e-4e90-fa9e-1b9eadc6705d"
      },
      "source": [
        "# plotting some random 10 images\n",
        "\n",
        "class_names = ['airplane','automobile','bird','cat','deer',\n",
        "               'dog','frog','horse','ship','truck']\n",
        "\n",
        "fig = plt.figure(figsize=(8,3))\n",
        "for i in range(num_classes):\n",
        "    ax = fig.add_subplot(2, 5, 1 + i, xticks=[], yticks=[])\n",
        "    idx = np.where(y_train[:]==i)[0]\n",
        "    features_idx = x_train[idx,::]\n",
        "    img_num = np.random.randint(features_idx.shape[0])\n",
        "    im = (features_idx[img_num,::])\n",
        "    ax.set_title(class_names[i])\n",
        "    plt.imshow(im)\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 576x216 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rAjNwK_BbF9v",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Convert class vectors to binary class matrices.\n",
        "y_train = keras.utils.to_categorical(y_train, num_classes)\n",
        "y_test = keras.utils.to_categorical(y_test, num_classes)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AZS20ZKzbF95",
        "colab_type": "code",
        "colab": {},
        "outputId": "7cf0325c-ff16-46ef-de90-1ed31c559fc0"
      },
      "source": [
        "# Printing sample data\n",
        "print(y_train[:10])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[[0. 0. 0. 0. 0. 0. 1. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]\n",
            " [0. 0. 0. 0. 1. 0. 0. 0. 0. 0.]\n",
            " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 1. 0. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 1. 0. 0.]\n",
            " [0. 0. 0. 0. 0. 0. 0. 0. 1. 0.]\n",
            " [0. 0. 0. 1. 0. 0. 0. 0. 0. 0.]]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "912A0AzzbF-F",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = Sequential()\n",
        "model.add(Conv2D(64, (3, 3), padding='same',\n",
        "                 input_shape=x_train.shape[1:]))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(64, (3, 3)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model.add(Dropout(0.25))\n",
        "\n",
        "model.add(Conv2D(128, (3, 3), padding='same'))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(Conv2D(128, (3, 3)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(BatchNormalization())\n",
        "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
        "model.add(Dropout(0.25))\n",
        "\n",
        "model.add(Flatten())\n",
        "model.add(Dense(512,kernel_regularizer=l2(0.01)))\n",
        "model.add(Activation('relu'))\n",
        "model.add(Dropout(0.5))\n",
        "model.add(Dense(num_classes))\n",
        "model.add(Activation('softmax'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2J9Y674XbF-R",
        "colab_type": "code",
        "colab": {},
        "outputId": "158c764e-fc95-4a50-ee77-70b7eedd3788"
      },
      "source": [
        "# summary of the model\n",
        "print(model.summary())"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_5 (Conv2D)            (None, 32, 32, 64)        1792      \n",
            "_________________________________________________________________\n",
            "activation_5 (Activation)    (None, 32, 32, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_5 (Batch (None, 32, 32, 64)        256       \n",
            "_________________________________________________________________\n",
            "conv2d_6 (Conv2D)            (None, 30, 30, 64)        36928     \n",
            "_________________________________________________________________\n",
            "activation_6 (Activation)    (None, 30, 30, 64)        0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_6 (Batch (None, 30, 30, 64)        256       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_3 (MaxPooling2 (None, 15, 15, 64)        0         \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 15, 15, 64)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_7 (Conv2D)            (None, 15, 15, 128)       73856     \n",
            "_________________________________________________________________\n",
            "activation_7 (Activation)    (None, 15, 15, 128)       0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_7 (Batch (None, 15, 15, 128)       512       \n",
            "_________________________________________________________________\n",
            "conv2d_8 (Conv2D)            (None, 13, 13, 128)       147584    \n",
            "_________________________________________________________________\n",
            "activation_8 (Activation)    (None, 13, 13, 128)       0         \n",
            "_________________________________________________________________\n",
            "batch_normalization_8 (Batch (None, 13, 13, 128)       512       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 6, 6, 128)         0         \n",
            "_________________________________________________________________\n",
            "dropout_4 (Dropout)          (None, 6, 6, 128)         0         \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 4608)              0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 512)               2359808   \n",
            "_________________________________________________________________\n",
            "activation_9 (Activation)    (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dropout_5 (Dropout)          (None, 512)               0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 10)                5130      \n",
            "_________________________________________________________________\n",
            "activation_10 (Activation)   (None, 10)                0         \n",
            "=================================================================\n",
            "Total params: 2,626,634\n",
            "Trainable params: 2,625,866\n",
            "Non-trainable params: 768\n",
            "_________________________________________________________________\n",
            "None\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MWOw59ABbF-b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# compile\n",
        "model.compile(loss='categorical_crossentropy',\n",
        "              optimizer='sgd',\n",
        "              metrics=['accuracy'])\n",
        "\n",
        "x_train = x_train.astype('float32')\n",
        "x_test = x_test.astype('float32')\n",
        "\n",
        "# Normalizing the input image\n",
        "x_train /= 255\n",
        "x_test /= 255\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GCtCUp22bF-n",
        "colab_type": "code",
        "colab": {},
        "outputId": "d61fe788-1d87-49c1-8e2f-2d6824ad5de8"
      },
      "source": [
        "# Training the model\n",
        "history = model.fit(x_train, y_train,\n",
        "              batch_size=batch_size,\n",
        "              epochs=epochs,\n",
        "              validation_data=(x_test, y_test),\n",
        "              shuffle=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 50000 samples, validate on 10000 samples\n",
            "Epoch 1/50\n",
            "50000/50000 [==============================] - 41s 827us/step - loss: 8.5320 - acc: 0.4192 - val_loss: 6.2763 - val_acc: 0.5320\n",
            "Epoch 2/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 4.9414 - acc: 0.5711 - val_loss: 3.7928 - val_acc: 0.6193\n",
            "Epoch 3/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 3.0706 - acc: 0.6477 - val_loss: 2.4057 - val_acc: 0.6991\n",
            "Epoch 4/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 2.0642 - acc: 0.6935 - val_loss: 1.7908 - val_acc: 0.6946\n",
            "Epoch 5/50\n",
            "50000/50000 [==============================] - 39s 788us/step - loss: 1.5183 - acc: 0.7242 - val_loss: 1.3366 - val_acc: 0.7413\n",
            "Epoch 6/50\n",
            "50000/50000 [==============================] - 39s 784us/step - loss: 1.2193 - acc: 0.7454 - val_loss: 1.1800 - val_acc: 0.7408\n",
            "Epoch 7/50\n",
            "50000/50000 [==============================] - 39s 784us/step - loss: 1.0453 - acc: 0.7661 - val_loss: 0.9936 - val_acc: 0.7743\n",
            "Epoch 8/50\n",
            "50000/50000 [==============================] - 39s 784us/step - loss: 0.9479 - acc: 0.7799 - val_loss: 1.1978 - val_acc: 0.6922\n",
            "Epoch 9/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.8848 - acc: 0.7924 - val_loss: 0.9494 - val_acc: 0.7726\n",
            "Epoch 10/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.8434 - acc: 0.8054 - val_loss: 0.9111 - val_acc: 0.7865\n",
            "Epoch 11/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.8124 - acc: 0.8168 - val_loss: 0.8809 - val_acc: 0.7972\n",
            "Epoch 12/50\n",
            "50000/50000 [==============================] - 39s 787us/step - loss: 0.7940 - acc: 0.8260 - val_loss: 0.8888 - val_acc: 0.7958\n",
            "Epoch 13/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.7748 - acc: 0.8359 - val_loss: 0.8692 - val_acc: 0.8063\n",
            "Epoch 14/50\n",
            "50000/50000 [==============================] - 39s 782us/step - loss: 0.7685 - acc: 0.8411 - val_loss: 0.9375 - val_acc: 0.7941\n",
            "Epoch 15/50\n",
            "50000/50000 [==============================] - 39s 782us/step - loss: 0.7610 - acc: 0.8497 - val_loss: 1.0530 - val_acc: 0.7629\n",
            "Epoch 16/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.7615 - acc: 0.8541 - val_loss: 0.9402 - val_acc: 0.8060\n",
            "Epoch 17/50\n",
            "50000/50000 [==============================] - 39s 782us/step - loss: 0.7551 - acc: 0.8625 - val_loss: 0.9706 - val_acc: 0.8045\n",
            "Epoch 18/50\n",
            "50000/50000 [==============================] - 39s 782us/step - loss: 0.7578 - acc: 0.8660 - val_loss: 0.9409 - val_acc: 0.8146\n",
            "Epoch 19/50\n",
            "50000/50000 [==============================] - 39s 783us/step - loss: 0.7507 - acc: 0.8708 - val_loss: 0.9106 - val_acc: 0.8249\n",
            "Epoch 20/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7614 - acc: 0.8739 - val_loss: 0.9214 - val_acc: 0.8296\n",
            "Epoch 21/50\n",
            "50000/50000 [==============================] - 39s 788us/step - loss: 0.7502 - acc: 0.8806 - val_loss: 0.9319 - val_acc: 0.8302\n",
            "Epoch 22/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7581 - acc: 0.8802 - val_loss: 1.0477 - val_acc: 0.7979\n",
            "Epoch 23/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7572 - acc: 0.8860 - val_loss: 0.9583 - val_acc: 0.8284\n",
            "Epoch 24/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7564 - acc: 0.8875 - val_loss: 0.9680 - val_acc: 0.8257\n",
            "Epoch 25/50\n",
            "50000/50000 [==============================] - 39s 789us/step - loss: 0.7522 - acc: 0.8897 - val_loss: 0.9740 - val_acc: 0.8268\n",
            "Epoch 26/50\n",
            "50000/50000 [==============================] - 40s 791us/step - loss: 0.7630 - acc: 0.8890 - val_loss: 0.9858 - val_acc: 0.8246\n",
            "Epoch 27/50\n",
            "50000/50000 [==============================] - 39s 788us/step - loss: 0.7557 - acc: 0.8954 - val_loss: 0.9905 - val_acc: 0.8280\n",
            "Epoch 28/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7510 - acc: 0.8980 - val_loss: 1.0122 - val_acc: 0.8262\n",
            "Epoch 29/50\n",
            "50000/50000 [==============================] - 40s 791us/step - loss: 0.7480 - acc: 0.8996 - val_loss: 0.9762 - val_acc: 0.8365\n",
            "Epoch 30/50\n",
            "50000/50000 [==============================] - 39s 784us/step - loss: 0.7474 - acc: 0.9008 - val_loss: 0.9970 - val_acc: 0.8352\n",
            "Epoch 31/50\n",
            "50000/50000 [==============================] - 39s 784us/step - loss: 0.7428 - acc: 0.9028 - val_loss: 1.0105 - val_acc: 0.8270\n",
            "Epoch 32/50\n",
            "50000/50000 [==============================] - 39s 787us/step - loss: 0.7466 - acc: 0.9036 - val_loss: 0.9937 - val_acc: 0.8391\n",
            "Epoch 33/50\n",
            "50000/50000 [==============================] - 39s 790us/step - loss: 0.7473 - acc: 0.9055 - val_loss: 1.0112 - val_acc: 0.8286\n",
            "Epoch 34/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7460 - acc: 0.9059 - val_loss: 1.0726 - val_acc: 0.8192\n",
            "Epoch 35/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7365 - acc: 0.9101 - val_loss: 1.0054 - val_acc: 0.8346\n",
            "Epoch 36/50\n",
            "50000/50000 [==============================] - 39s 787us/step - loss: 0.7373 - acc: 0.9100 - val_loss: 1.0239 - val_acc: 0.8360\n",
            "Epoch 37/50\n",
            "50000/50000 [==============================] - 39s 787us/step - loss: 0.7341 - acc: 0.9129 - val_loss: 1.0220 - val_acc: 0.8337\n",
            "Epoch 38/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7324 - acc: 0.9128 - val_loss: 1.0448 - val_acc: 0.8318\n",
            "Epoch 39/50\n",
            "50000/50000 [==============================] - 39s 786us/step - loss: 0.7276 - acc: 0.9158 - val_loss: 1.0725 - val_acc: 0.8242\n",
            "Epoch 40/50\n",
            "50000/50000 [==============================] - 39s 787us/step - loss: 0.7318 - acc: 0.9139 - val_loss: 1.0940 - val_acc: 0.8173\n",
            "Epoch 41/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7189 - acc: 0.9191 - val_loss: 1.0308 - val_acc: 0.8370\n",
            "Epoch 42/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7233 - acc: 0.9166 - val_loss: 1.0374 - val_acc: 0.8326\n",
            "Epoch 43/50\n",
            "50000/50000 [==============================] - 39s 785us/step - loss: 0.7124 - acc: 0.9201 - val_loss: 1.0788 - val_acc: 0.8247\n",
            "Epoch 44/50\n",
            "50000/50000 [==============================] - 40s 793us/step - loss: 0.7153 - acc: 0.9184 - val_loss: 1.0054 - val_acc: 0.8435\n",
            "Epoch 45/50\n",
            "50000/50000 [==============================] - 40s 792us/step - loss: 0.7073 - acc: 0.9221 - val_loss: 1.1571 - val_acc: 0.7981\n",
            "Epoch 46/50\n",
            "50000/50000 [==============================] - 40s 792us/step - loss: 0.7095 - acc: 0.9211 - val_loss: 1.0430 - val_acc: 0.8284\n",
            "Epoch 47/50\n",
            "50000/50000 [==============================] - 40s 791us/step - loss: 0.7146 - acc: 0.9205 - val_loss: 1.0360 - val_acc: 0.8346\n",
            "Epoch 48/50\n",
            "50000/50000 [==============================] - 40s 790us/step - loss: 0.7083 - acc: 0.9228 - val_loss: 1.0431 - val_acc: 0.8347\n",
            "Epoch 49/50\n",
            "50000/50000 [==============================] - 40s 792us/step - loss: 0.6993 - acc: 0.9241 - val_loss: 1.0457 - val_acc: 0.8320\n",
            "Epoch 50/50\n",
            "50000/50000 [==============================] - 39s 788us/step - loss: 0.6966 - acc: 0.9244 - val_loss: 1.0214 - val_acc: 0.8406\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nw3PjlaVbF-x",
        "colab_type": "code",
        "colab": {},
        "outputId": "66e4de52-bc85-4ac8-a744-b0c8d10aefeb"
      },
      "source": [
        "# list all data in history\n",
        "print(history.history.keys())\n",
        "# summarize history for accuracy\n",
        "plt.plot(history.history['acc'])\n",
        "plt.plot(history.history['val_acc'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history.history['loss'])\n",
        "plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "dict_keys(['val_loss', 'val_acc', 'loss', 'acc'])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CWq_O_bCbF-4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}