convnet-mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "computer-vision-convnets",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "[View in Colaboratory](https://colab.research.google.com/github/silentvoice/dl-with-keras/blob/master/computer_vision_convnets.ipynb)"
      ]
    },
    {
      "metadata": {
        "id": "rCxHTcOHpnjj",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Computer Vision using Convolutional Neural Networks"
      ]
    },
    {
      "metadata": {
        "id": "2UlCGg0IpzTW",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Check Tensorflow Version"
      ]
    },
    {
      "metadata": {
        "id": "GN9C5Hh3pXS0",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "import tensorflow as tf"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "hz8_FKF1nBZM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "5da2aaad-ff91-469c-9d62-eae9e16ba0a3"
      },
      "cell_type": "code",
      "source": [
        "tf.test.gpu_device_name()"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'/device:GPU:0'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "metadata": {
        "id": "NBO8e2kHoFSI",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "2ce3cdee-e253-4052-cf42-6149c261adc7"
      },
      "cell_type": "code",
      "source": [
        "tf.__version__"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'1.9.0-rc2'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "metadata": {
        "id": "JFlI4hFYpNBo",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Imports"
      ]
    },
    {
      "metadata": {
        "id": "QEIfV4L5qMZP",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from tensorflow import keras\n",
        "from tensorflow.python.keras.datasets import mnist\n",
        "from tensorflow.python.keras.utils import to_categorical\n",
        "from tensorflow.python.keras import layers, models\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FX1hYnllqQDI",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Load Mnist Dataset"
      ]
    },
    {
      "metadata": {
        "id": "FWLu6M9bqxcx",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "(train_data, train_labels), (test_data, test_labels) = mnist.load_data()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "y0Cox0a0rxf2",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "33b9520f-f506-447e-a9a3-691c6303d2d9"
      },
      "cell_type": "code",
      "source": [
        "train_data.shape"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 36
        }
      ]
    },
    {
      "metadata": {
        "id": "2qZgCLo_sK52",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "dda2c97d-ed5b-4538-a8ed-39bcf2346a1d"
      },
      "cell_type": "code",
      "source": [
        "train_labels.shape"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 37
        }
      ]
    },
    {
      "metadata": {
        "id": "TqLtK0r6fe9b",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "4fdee97c-894f-4b33-8e0f-2cf011f620bb"
      },
      "cell_type": "code",
      "source": [
        "test_data.shape"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000, 28, 28)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "metadata": {
        "id": "JI26BqkP0Qwg",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Data Preparation"
      ]
    },
    {
      "metadata": {
        "id": "J0OPDlT-gQlh",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "### Reshape and scale the train and test images"
      ]
    },
    {
      "metadata": {
        "id": "d1FshZ5u0eWY",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "train_images = train_data.reshape((60000, 28, 28, 1)).astype('float32') / 255\n",
        "test_images = test_data.reshape((10000, 28, 28, 1)).astype('float32') / 255"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "fxN_Lngl1fqz",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "f038dde6-9e1e-42ba-f820-7c2c4addd57f"
      },
      "cell_type": "code",
      "source": [
        "train_images.shape"
      ],
      "execution_count": 40,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28, 1)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 40
        }
      ]
    },
    {
      "metadata": {
        "id": "83Icrv904CPo",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "### Onehot encode the train and test labels"
      ]
    },
    {
      "metadata": {
        "id": "jqXSq_hc5eHp",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "train_labels = to_categorical(train_labels)\n",
        "test_labels = to_categorical(test_labels)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "8yH3NVPi5XWZ",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Model"
      ]
    },
    {
      "metadata": {
        "id": "4BbqBG3cgkPc",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def create_model():\n",
        "  model = models.Sequential()\n",
        "  model.add(layers.Conv2D(32, (3,3), activation='relu', input_shape=(28, 28, 1)))\n",
        "  model.add(layers.MaxPooling2D(2,2))\n",
        "  model.add(layers.Conv2D(64, (3,3), activation='relu'))\n",
        "  model.add(layers.MaxPooling2D(2,2))\n",
        "  model.add(layers.Conv2D(64, (3,3), activation='relu'))\n",
        "  model.add(layers.Flatten())\n",
        "  model.add(layers.Dense(64, activation='relu'))\n",
        "  model.add(layers.Dense(10, activation='softmax'))\n",
        "  model.compile(optimizer='rmsprop',\n",
        "                loss='categorical_crossentropy',\n",
        "                metrics=['accuracy'])\n",
        "  return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "X8DuRHSl7Kbk",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model = create_model()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QL36FeP-7Y6g",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 408
        },
        "outputId": "7e862fee-22f0-404a-8a3e-45afe113b23e"
      },
      "cell_type": "code",
      "source": [
        "model.summary()"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_6 (Conv2D)            (None, 26, 26, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d_4 (MaxPooling2 (None, 13, 13, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_7 (Conv2D)            (None, 11, 11, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_5 (MaxPooling2 (None, 5, 5, 64)          0         \n",
            "_________________________________________________________________\n",
            "conv2d_8 (Conv2D)            (None, 3, 3, 64)          36928     \n",
            "_________________________________________________________________\n",
            "flatten_2 (Flatten)          (None, 576)               0         \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 64)                36928     \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 10)                650       \n",
            "=================================================================\n",
            "Total params: 93,322\n",
            "Trainable params: 93,322\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "DuSQnWXf7m4S",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Train the Model"
      ]
    },
    {
      "metadata": {
        "id": "3PPbdUhx7YMT",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 224
        },
        "outputId": "f67d7377-61ca-49e0-cdb6-995a6a42f6cc"
      },
      "cell_type": "code",
      "source": [
        "model.fit(train_images, train_labels, epochs=5, batch_size=64)"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "60000/60000 [==============================] - 11s 175us/step - loss: 0.1746 - acc: 0.9456\n",
            "Epoch 2/5\n",
            "60000/60000 [==============================] - 11s 179us/step - loss: 0.0484 - acc: 0.9848\n",
            "Epoch 3/5\n",
            "47744/60000 [======================>.......] - ETA: 2s - loss: 0.0330 - acc: 0.9894"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "60000/60000 [==============================] - 11s 178us/step - loss: 0.0339 - acc: 0.9892\n",
            "Epoch 4/5\n",
            "60000/60000 [==============================] - 10s 174us/step - loss: 0.0253 - acc: 0.9920\n",
            "Epoch 5/5\n",
            "60000/60000 [==============================] - 11s 182us/step - loss: 0.0201 - acc: 0.9940\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fb135142588>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 45
        }
      ]
    },
    {
      "metadata": {
        "id": "qaoX1DHtTqkI",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Evaluate the model\n",
        "\n"
      ]
    },
    {
      "metadata": {
        "id": "XPtHfaufSixn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "a8895a07-5682-4ebb-a58e-bd0f0f1bfb87"
      },
      "cell_type": "code",
      "source": [
        "test_loss, test_accuracy = model.evaluate(test_images, test_labels)"
      ],
      "execution_count": 47,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "10000/10000 [==============================] - 1s 69us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "Svl5rzigUHJ9",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "fad724b9-6e1c-41d3-d369-46f0e3849703"
      },
      "cell_type": "code",
      "source": [
        "test_accuracy"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.9909"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "metadata": {
        "id": "GiEeNUqsucux",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "ad624297-e98d-40a9-ded0-053e6738440c"
      },
      "cell_type": "code",
      "source": [
        "test_loss"
      ],
      "execution_count": 49,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "0.030745380283371922"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "metadata": {
        "id": "wk2y8ymf2pjo",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}