Denoising Autoencoder_MNIST.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Denoising Autoencoder_MNIST.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "pv6Y4BEnOiNU",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "5561bc69-36c2-49eb-d8fc-fecfb0593ebd"
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from keras.models import Model\n",
        "from keras.layers import Input, Conv2D, MaxPooling2D\n",
        "from keras.layers import UpSampling2D, LeakyReLU"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-11iCvfoPiY4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "## Load data\n",
        "from keras.datasets import mnist"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sNVAdLnVPk9J",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "## Input image dimensions\n",
        "img_rows, img_cols = 28, 28\n",
        "input_shape = (img_rows, img_cols, 1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "k6w8K8tmPq82",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        },
        "outputId": "cd3b2624-a36a-4b24-8482-e5482db6e77e"
      },
      "source": [
        "## The data, shuffled and split between train and test sets\n",
        "(X_train, _), (X_test, _) = mnist.load_data()\n",
        "\n",
        "X_train = X_train.reshape(X_train.shape[0], img_rows, img_cols, 1)\n",
        "X_test = X_test.reshape(X_test.shape[0], img_rows, img_cols, 1)\n",
        "\n",
        "X_train = X_train.astype('float32') / 255.\n",
        "X_test = X_test.astype('float32') / 255."
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading data from https://s3.amazonaws.com/img-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 2s 0us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CwsLznZGQKAz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "noise_factor = 0.5\n",
        "\n",
        "X_train_noisy = X_train + noise_factor * np.random.normal(loc = 0.0, scale = 1.0, \n",
        "                                                          size = X_train.shape)\n",
        "X_test_noisy = X_test + noise_factor * np.random.normal(loc = 0.0, scale = 1.0, \n",
        "                                                        size = X_test.shape)\n",
        "\n",
        "X_train_noisy = np.clip(X_train_noisy, 0., 1.)\n",
        "X_test_noisy = np.clip(X_test_noisy, 0., 1.)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C-oabIC_Qu1P",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 53
        },
        "outputId": "47f127cc-b5c3-43a2-d1aa-fff0e7e8ab8a"
      },
      "source": [
        "print(X_train.shape[0], ' training samples')\n",
        "print(X_test.shape[0], ' test samples')"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "60000  training samples\n",
            "10000  test samples\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IHTBT6Q4Q2AR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def DAE_CNN(features_shape, act = 'relu'):\n",
        "  \n",
        "  #Input\n",
        "  x = Input(name = 'inputs', shape = features_shape, dtype = 'float32')\n",
        "  o = x\n",
        "  \n",
        "  \n",
        "  #Encoder\n",
        "  o = Conv2D(32, (3, 3), activation = act, padding = 'same',\n",
        "             strides = (1, 1), name = 'en_conv1')(o)\n",
        "  o = MaxPooling2D((2, 2), strides = (2, 2), padding = 'same',\n",
        "                   name = 'en_pool1')(o)\n",
        "  o = Conv2D(32, (3, 3), activation = act, padding = 'same',\n",
        "             strides = (1, 1), name = 'en_conv2')(o)\n",
        "  enc = MaxPooling2D((2, 2), strides = (2, 2), padding = 'same',\n",
        "                     name = 'en_pool2')(o)\n",
        "  \n",
        "  \n",
        "  #Decoder\n",
        "  o = Conv2D(32, (3, 3), activation = act, padding = 'same',\n",
        "             strides = (1, 1), name = 'de_conv1')(enc)\n",
        "  o = UpSampling2D((2, 2), name = 'upsampling1')(o)\n",
        "  o = Conv2D(32, (3, 3), activation = act, padding = 'same',\n",
        "             strides = (1, 1), name = 'de_conv2')(o)\n",
        "  o = UpSampling2D((2, 2), name = 'upsampling2')(o)\n",
        "  dec = Conv2D(1, (3, 3), activation = 'sigmoid', padding = 'same',\n",
        "               strides = (1, 1), name = 'de_conv3')(o)\n",
        "  \n",
        "  \n",
        "  #Print network summary\n",
        "  Model(inputs = x, outputs = dec).summary()\n",
        "  \n",
        "  return Model(inputs = x, outputs = dec)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ydbE0GGaTifn",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "39805431-1a26-4f70-9327-cf3ac2cf0b08"
      },
      "source": [
        "batch_size = 128\n",
        "epochs = 40\n",
        "\n",
        "autoencoder = DAE_CNN(input_shape, act = LeakyReLU(alpha = 0.1))\n",
        "autoencoder.compile(optimizer = 'adadelta', \n",
        "                    loss = 'binary_crossentropy')\n",
        "\n",
        "autoencoder.fit(X_train_noisy, X_train,\n",
        "                epochs = epochs,\n",
        "                batch_size = batch_size,\n",
        "                shuffle = True,\n",
        "                validation_data = (X_test_noisy, X_test))\n",
        "\n",
        "decoded_imgs = autoencoder.predict(X_test_noisy)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0902 19:14:11.957674 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
            "\n",
            "W0902 19:14:11.998645 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
            "\n",
            "/usr/local/lib/python3.6/dist-packages/keras/activations.py:235: UserWarning: Do not pass a layer instance (such as LeakyReLU) as the activation argument of another layer. Instead, advanced activation layers should be used just like any other layer in a model.\n",
            "  identifier=identifier.__class__.__name__))\n",
            "W0902 19:14:12.005703 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
            "\n",
            "W0902 19:14:12.032094 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4267: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.\n",
            "\n",
            "W0902 19:14:12.060331 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:2239: The name tf.image.resize_nearest_neighbor is deprecated. Please use tf.compat.v1.image.resize_nearest_neighbor instead.\n",
            "\n",
            "W0902 19:14:12.091259 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
            "\n",
            "W0902 19:14:12.105010 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3657: The name tf.log is deprecated. Please use tf.math.log instead.\n",
            "\n",
            "W0902 19:14:12.110268 139940059129728 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/nn_impl.py:180: add_dispatch_support.<locals>.wrapper (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Model: \"model_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "inputs (InputLayer)          (None, 28, 28, 1)         0         \n",
            "_________________________________________________________________\n",
            "en_conv1 (Conv2D)            (None, 28, 28, 32)        320       \n",
            "_________________________________________________________________\n",
            "en_pool1 (MaxPooling2D)      (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "en_conv2 (Conv2D)            (None, 14, 14, 32)        9248      \n",
            "_________________________________________________________________\n",
            "en_pool2 (MaxPooling2D)      (None, 7, 7, 32)          0         \n",
            "_________________________________________________________________\n",
            "de_conv1 (Conv2D)            (None, 7, 7, 32)          9248      \n",
            "_________________________________________________________________\n",
            "upsampling1 (UpSampling2D)   (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "de_conv2 (Conv2D)            (None, 14, 14, 32)        9248      \n",
            "_________________________________________________________________\n",
            "upsampling2 (UpSampling2D)   (None, 28, 28, 32)        0         \n",
            "_________________________________________________________________\n",
            "de_conv3 (Conv2D)            (None, 28, 28, 1)         289       \n",
            "=================================================================\n",
            "Total params: 28,353\n",
            "Trainable params: 28,353\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "W0902 19:14:12.452394 139940059129728 deprecation_wrapper.py:119] From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:1033: The name tf.assign_add is deprecated. Please use tf.compat.v1.assign_add instead.\n",
            "\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Train on 60000 samples, validate on 10000 samples\n",
            "Epoch 1/40\n",
            "60000/60000 [==============================] - 10s 173us/step - loss: 0.1725 - val_loss: 0.1222\n",
            "Epoch 2/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.1175 - val_loss: 0.1102\n",
            "Epoch 3/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.1101 - val_loss: 0.1064\n",
            "Epoch 4/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.1065 - val_loss: 0.1054\n",
            "Epoch 5/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.1044 - val_loss: 0.1040\n",
            "Epoch 6/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.1031 - val_loss: 0.1037\n",
            "Epoch 7/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.1021 - val_loss: 0.0999\n",
            "Epoch 8/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.1014 - val_loss: 0.1004\n",
            "Epoch 9/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.1008 - val_loss: 0.1000\n",
            "Epoch 10/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.1004 - val_loss: 0.0990\n",
            "Epoch 11/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0998 - val_loss: 0.0999\n",
            "Epoch 12/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0996 - val_loss: 0.0992\n",
            "Epoch 13/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0993 - val_loss: 0.0987\n",
            "Epoch 14/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0990 - val_loss: 0.0978\n",
            "Epoch 15/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0988 - val_loss: 0.0980\n",
            "Epoch 16/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0986 - val_loss: 0.0977\n",
            "Epoch 17/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.0983 - val_loss: 0.0972\n",
            "Epoch 18/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.0981 - val_loss: 0.0970\n",
            "Epoch 19/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0978 - val_loss: 0.0973\n",
            "Epoch 20/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0978 - val_loss: 0.0969\n",
            "Epoch 21/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.0976 - val_loss: 0.0976\n",
            "Epoch 22/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0974 - val_loss: 0.0968\n",
            "Epoch 23/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0973 - val_loss: 0.0969\n",
            "Epoch 24/40\n",
            "60000/60000 [==============================] - 4s 64us/step - loss: 0.0971 - val_loss: 0.0968\n",
            "Epoch 25/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0971 - val_loss: 0.0984\n",
            "Epoch 26/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0969 - val_loss: 0.0965\n",
            "Epoch 27/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0969 - val_loss: 0.0959\n",
            "Epoch 28/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0967 - val_loss: 0.0963\n",
            "Epoch 29/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0965 - val_loss: 0.0958\n",
            "Epoch 30/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0965 - val_loss: 0.0956\n",
            "Epoch 31/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0964 - val_loss: 0.0965\n",
            "Epoch 32/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0963 - val_loss: 0.0960\n",
            "Epoch 33/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0962 - val_loss: 0.0956\n",
            "Epoch 34/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0962 - val_loss: 0.0966\n",
            "Epoch 35/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0961 - val_loss: 0.0956\n",
            "Epoch 36/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0959 - val_loss: 0.0956\n",
            "Epoch 37/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0959 - val_loss: 0.0951\n",
            "Epoch 38/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0958 - val_loss: 0.0958\n",
            "Epoch 39/40\n",
            "60000/60000 [==============================] - 4s 65us/step - loss: 0.0958 - val_loss: 0.0951\n",
            "Epoch 40/40\n",
            "60000/60000 [==============================] - 4s 66us/step - loss: 0.0957 - val_loss: 0.0960\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "42ddKni_VRH3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 264
        },
        "outputId": "ec4af252-c813-4624-f262-1e3e8994ad14"
      },
      "source": [
        "n = 10\n",
        "\n",
        "plt.figure(figsize = (20, 4))\n",
        "\n",
        "for i in range(1, n + 1):\n",
        "  ##Displaying original images\n",
        "  ax = plt.subplot(2, n, i)\n",
        "  plt.imshow(X_test_noisy[i].reshape(28, 28))\n",
        "  plt.gray()\n",
        "  ax.get_xaxis().set_visible(False)\n",
        "  ax.get_yaxis().set_visible(False)\n",
        "  \n",
        "  ##Displaying reconstructed images\n",
        "  ax = plt.subplot(2, n, i + n)\n",
        "  plt.imshow(decoded_imgs[i].reshape(28, 28))\n",
        "  plt.gray()\n",
        "  ax.get_xaxis().set_visible(False)\n",
        "  ax.get_yaxis().set_visible(False)\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABGoAAADjCAYAAADdR/IFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 1440x288 with 20 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}