Autoencoder(from_noise_to_original).ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Autoencoder(from noise to original).ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOSvpH/FTPbrn/+T4y9Dd+m",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Neel7317/Generative-Models/blob/main/Autoencoder(from_noise_to_original).ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kld1OggOoCV9",
        "outputId": "93ee89fa-22f1-424a-ada3-8edfcc8d03fb"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 0s 0us/step\n",
            "11501568/11490434 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "from keras.datasets import mnist\n",
        "import tensorflow as tf\n",
        "### Importing Libraries\n",
        "\n",
        "import keras\n",
        "from keras import callbacks\n",
        "from keras.models import Model\n",
        "from tensorflow.keras.optimizers import Adadelta\n",
        "from keras.layers import Input, Conv2D, MaxPool2D, UpSampling2D\n",
        "\n",
        "### Downloading and Preprocessing of dataset and adding some noise to it.\n",
        "\n",
        "import numpy as np\n",
        "\n",
        "(trainX, trainy), (testX, testy) = mnist.load_data()\n",
        "\n",
        "# to convert values from 0 to 255 into range 0 to 1.\n",
        "trainX = np.expand_dims(trainX, axis=-1)\n",
        "testX = np.expand_dims(testX, axis=-1)\n",
        "trainX = trainX.astype(\"float32\") / 255.0\n",
        "testX = testX.astype(\"float32\") / 255.0"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "RUC-k6Xdp4jD"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#add noise to the images\n",
        "trainNoise = np.random.normal(loc=0.5, scale=0.5, size=trainX.shape)\n",
        "testNoise = np.random.normal(loc=0.5, scale=0.5, size=testX.shape)\n",
        "trainXNoisy = np.clip(trainX + trainNoise, 0, 1)\n",
        "testXNoisy = np.clip(testX + testNoise, 0, 1)"
      ],
      "metadata": {
        "id": "-uN_-Swwp4O6"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.figure(figsize=(10, 10))\n",
        "for i in range(5):\n",
        "  plt.subplot(1, 5, i+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.grid(False)\n",
        "  plt.imshow(trainX[i].reshape(28, 28), cmap='gray')\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "\n",
        "plt.figure(figsize=(10, 10))\n",
        "for i in range(5):\n",
        "  plt.subplot(1, 5, i+1)\n",
        "  plt.xticks([])\n",
        "  plt.yticks([])\n",
        "  plt.grid(False)\n",
        "  plt.imshow(trainXNoisy[i].reshape(28, 28), cmap='gray')\n",
        "plt.tight_layout()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 307
        },
        "id": "OdoaGOAPpTv9",
        "outputId": "1453e26c-79c2-4613-8eb1-3bbc8c7201ab"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 5 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAACRCAYAAADetU5gAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAQ+UlEQVR4nO3de2xUxRfA8btAKCgpFYqIEigq1Kgp1fKSEIpS0CgqUBUboIAGiAg0RggRK8FoEXmYUAQlNIA8EiBiKWAIGMtDFBtKhQQrWB+hKTS1gKW0IA2yvz9+v18yM2d0dy+73d3u9/PfOTn33oGO9Hgzc8fj9XodAAAAAP/VKtwDAAAAACIJDTIAAACgoEEGAAAAFDTIAAAAgIIGGQAAAFDQIAMAAACKNoEUezwevgkXw7xer8fNdcyb2OZm3jBnYt4Fr9fbJdCLmDcxj3kDN6zzhjfIAIBIczbcA0BUYt7ADeu8oUEGAAAAFDTIAAAAgIIGGQAAAFDQIAMAAAAKGmQAAABAQYMMAAAAKGiQAQAAAAUNMgAAAKCgQQYAAAAUNMgAAACAggYZAAAAUNAgAwAAAAoaZAAAAEBBgwwAAAAoaJABAAAARZtwDwCIRWlpaVo8c+ZMUZOdna3FGzduFDUrV67U4rKysiCMDgCA2MYbZAAAAEBBgwwAAAAoaJABAAAABQ0yAAAAoPB4vV7/iz0e/4ujTOvWrbW4Y8eOru5j22x12223aXFycrKoef3117V42bJloiYrK0vk/vrrLy1evHixqHn33Xftgw2Q1+v1uLmuJc8bf6SmpopccXGxFsfHx7u69+XLl7W4c+fOru4TSm7mTazPmeY0fPhwkduyZYvIpaena/GZM2dCNibHcY57vd5+gV7EvAmO3NxckTN/j7RqJd+vDRs2TOQOHToUtHH5gXkDN6zzhjfIAAAAgIIGGQAAAFDQIAMAAACKqD8opEePHlrctm1bUTN48GAtHjJkiKhJSEjQ4szMzCCMzq6qqkrk8vPztXjMmDGi5sqVKyJ38uRJLW7m9V4wDBgwQOR27NghcuYad9teAPPn3dTUJGrMNceDBg0SNbbDQ2z3aomGDh0qcubfWWFhYXMNJyL1799f5I4dOxaGkSBcJk+erMXz5s0TNTdv3vR5n0D2NAGRjjfIAAAAgIIGGQAAAFDQIAMAAAAKGmQAAABAEVWb9Pw5cMHtAR+hZG5usH2EvaGhQYttH+qvrq4WuT///FOLQ/zx/phmHvjiOI7z6KOPavHmzZtFTbdu3Vw9r6KiQouXLFkiarZu3arF3377raixzbcPPvjA1Ziije3ggt69e2txrG3SMw946NWrl6jp2bOnyHk8rs4JQhQwf97t2rUL00gQbAMHDtTiCRMmiBrzECDHcZyHHnrI573nzJmjxefPnxc15kcRbL8jS0pKfD4rHHiDDAAAAChokAEAAAAFDTIAAACgoEEGAAAAFFG1Sa+yslLkLl68qMWh3KRnW0heV1enxY8//rioMU8t27RpU3AHhmaxZs0akcvKygrZ88wNgB06dBA15smJtk1pKSkpQR1XNMnOzha5o0ePhmEkkcPcNDp16lRRY9tIc/r06ZCNCc0nIyND5GbNmuXzOvPnP2rUKFFTU1PjfmC4ZePGjRO5FStWaHFiYqKosW3APXjwoBZ36dJF1CxdutTnmMx72+7z8ssv+7xPOPAGGQAAAFDQIAMAAAAKGmQAAABAEVVrkC9duiRyc+fO1WLbuqgffvhBi/Pz830+68SJEyI3YsQIkWtsbNRi28e1c3JyfD4PkSctLU2Ln3nmGVHjz+EJ5jphx3Gc3bt3a/GyZctEjfnRdXMeO448KOaJJ55wNcaWyjwUA45TUFDgs8Y8pAbRyzyoYf369aLGn7075nrTs2fP3trAEJA2bWS71q9fPy1eu3atqDEPuDp8+LCoee+990TuyJEjWhwXFydqtm/frsUjR44UNabS0lKfNZGC3x4AAACAggYZAAAAUNAgAwAAAAoaZAAAAEARVZv0bHbu3KnFxcXFoubKlSta3LdvX1Hz6quvarFt05S5Ic/mxx9/FLlp06b5vA7hlZqaKnJfffWVFsfHx4sar9erxXv37hU1tsNE0tPTtTg3N1fUmJupamtrRc3Jkye1+ObNm6LGtrnQPISkrKxM1EQj81CUrl27hmkkkcufDVnm3Ef0mjRpkhbffffdPq8xD4lwHMfZuHFjsIYEFyZMmCBy/my4Nf9bth0mUl9f7/M+tuv82ZRXVVWlxZ999pnPayIFb5ABAAAABQ0yAAAAoKBBBgAAABRRvwbZ5M9amsuXL/usmTp1qsht27ZN5GxrPhH5+vTpo8XmgTOOI9dqXrhwQdRUV1drsW19VUNDg8h9+eWX/xoHU/v27UXuzTff1OLx48eH7PnN6emnn9Zi2589ltjWYPfq1cvndefOnQvFcBBiiYmJIvfKK69ose13Vl1dnRa///77wR0YAmYe3jF//nxRY+6BWb16tagx97f40yPZvP32266umz17thbb9tJEKt4gAwAAAAoaZAAAAEBBgwwAAAAoaJABAAAARYvbpOePhQsXilxaWpoWmwc5OI7jZGRkiNz+/fuDNi6ERlxcnMiZB8GYm7scRx4wk52dLWpKS0u1OFo2hfXo0SPcQwiJ5ORknzW2w3xaKtuBR+bGvZ9//lnUmHMfkScpKUnkduzY4epeK1eu1OIDBw64ug/cWbBggciZm/KamppEzb59+7R43rx5oubatWs+n9+uXTuRMw8Bsf3O8Hg8Wmzb3FlUVOTz+ZGKN8gAAACAggYZAAAAUNAgAwAAAIqYXIPc2NgocubBIGVlZaJm7dq1Imeu1TLXpDqO46xatUqLzY97I7QeeeQRkbOtOTY9//zzWnzo0KGgjQnhc+zYsXAPwZX4+Hgtfuqpp0TNhAkTtNhcR2hjHkjgOPLgCEQe288/JSXF53Vff/21yK1YsSIoY4JvCQkJIjdjxgyRM/sEc72x4zjO6NGjA37+/fffL3JbtmwROXNfls3nn3+uxUuWLAl4PJGMN8gAAACAggYZAAAAUNAgAwAAAAoaZAAAAEARk5v0bH799Vctnjx5sqhZv369yE2cOPFfY8dxnNtvv12LN27cKGqqq6v9GSZc+Oijj0TO/MC5bQNeNG7Ka9VK/j/vzZs3wzCSyNWpU6eg3Kdv374iZ84r2+FC3bt31+K2bduKmvHjx4uc+bO1HQBQUlKixdevXxc1bdro/+wfP35c1CDymBuyFi9e7Nd1R44c0eJJkyaJmsuXL7sfGAJi++89MTHR53WzZ88WuTvvvFOLp0yZImqee+45LX744YdFTYcOHUTO3CRo+7jA5s2btdj2AYRoxhtkAAAAQEGDDAAAAChokAEAAAAFDTIAAACgYJPePygsLBS5iooKkTM3gA0fPlzULFq0SIt79uwpavLy8rT43Llzfo0TulGjRolcamqqyJkbDnbt2hWyMTUn24Y82+aKEydONMdwmp25cc32Z//000+1eP78+a6eZTu1zNykd+PGDVFz9epVLS4vLxc169atEznzlE7bJtKamhotrqqqEjXt27fX4tOnT4sahFdSUpLI7dixw9W9fvvtNy025wiaV1NTk8jV1taKXJcuXbT4999/FzVuTuU9f/68yNXX14tct27dtPjChQuiZvfu3QE/P5rwBhkAAABQ0CADAAAAChpkAAAAQMEa5ACcOnVK5F566SUtfvbZZ0WNecDI9OnTRU3v3r21eMSIEW6GGPPM9ZWOY/8w+x9//KHF27ZtC9mYgiUuLk7kFi5c6PO64uJikXvrrbeCMaSIM2PGDC0+e/asqBk8eHBQnlVZWSlyO3fu1OKffvpJ1Hz//fdBeb7NtGnTtNhcx+g4ck0qIs+8efNEzu2BP/4eKILmUVdXJ3LmITCO4zh79uzRYtsBR+YBZ0VFRaJmw4YNWnzp0iVRs3XrVpEz1yDbalo63iADAAAAChpkAAAAQEGDDAAAAChokAEAAAAFm/RukbngftOmTaKmoKBAi9u0kX/tQ4cO1eJhw4aJmoMHDwY+QFhdv35di6urq8M0kn9mbsrLzc0VNXPnztVi28EQy5cvF7mGhoZbHF10+PDDD8M9hGZlO6jI5PbACYSOeZjRyJEjXd3HtknrzJkzru6F5lNSUiJytg22wWD2Go7jOOnp6SJnbgqNxc29vEEGAAAAFDTIAAAAgIIGGQAAAFCwBjkAKSkpIvfCCy9ocf/+/UWNbc2xqby8XIsPHz4c4OgQiF27doV7CBpzDaLjyPXF48aNEzXmmsPMzMzgDgwtTmFhYbiHAMP+/fu1+I477vB5je3AmcmTJwdrSGihbIdp2Q6h8Xq9WsxBIQAAAECMo0EGAAAAFDTIAAAAgIIGGQAAAFCwSe9/kpOTtXjmzJmiZuzYsSJ31113Bfysv//+W+TMgypsi+bhm8fj8Ss3evRoLc7JyQnZmGzeeOMNLX7nnXdETceOHbV4y5YtoiY7Ozu4AwPQ7Dp37qzF/vz7v3r1apGLlQOA4N6+ffvCPYSowRtkAAAAQEGDDAAAAChokAEAAABFTKxBNtcJZ2VliRpzzXFSUlLQnl9aWqrFeXl5oibSDq6IVubHzf8pZ86J/Px8UbNu3TotvnjxoqgZNGiQFk+cOFHU9O3bV+S6d++uxZWVlaLGXCtmW3MI/Bvb+vs+ffpose3ACYTO+vXrRa5Vq8DfVX333XfBGA5izJNPPhnuIUQN3iADAAAAChpkAAAAQEGDDAAAAChokAEAAABF1G/S69q1qxY/+OCDoubjjz/W4gceeCBozy8pKdHipUuXipqioiIt5hCQ8GvdurUWz5gxQ9RkZmZqcX19vajp3bu3q+ebG2wOHDggahYsWODq3sD/2TaoutkQBvdSU1O1OCMjQ9SYvxOamppEzapVq7S4pqYmCKNDrLn33nvDPYSowb+UAAAAgIIGGQAAAFDQIAMAAAAKGmQAAABAEbGb9Dp16iRya9asETlzA0SwFqDbTilavny5yJmnnV27di0oz4c7R48eFbljx46JXP/+/X3eyzxtz9wQamM7bW/r1q0il5OT4/NeQCg89thjWrxhw4bwDCRGJCQkaLH574rNuXPnRG7OnDlBGxNi1zfffCNyto27fEyAN8gAAACAhgYZAAAAUNAgAwAAAIqwrEEeOHCgyM2dO1eLBwwYIGruueeeoDz/6tWrIpefn6/FixYtEjWNjY1BeT5Cp6qqSuTGjh0rctOnT9fi3NxcV89bsWKFFn/yySei5pdffnF1b+BWeTyecA8BQAQ5deqUyFVUVIicuZ/rvvvuEzW1tbXBG1gE4g0yAAAAoKBBBgAAABQ0yAAAAICCBhkAAABQhGWT3pgxY/zK+aO8vFyL9+zZI2pu3LihxbYDP+rq6lw9H5Gvurpa5BYuXPivMRCN9u7dq8UvvvhimEaC/zt9+rQW2w6hGjJkSHMNBxBsHyUoKCjQ4ry8PFEza9YsLTb7sWjHG2QAAABAQYMMAAAAKGiQAQAAAIXH6/X6X+zx+F+MFsfr9bo6dYB5E9vczBvmTMw77vV6+wV6EfMm5jFvXIiPjxe57du3a3FGRoao+eKLL7R4ypQpoiZKDlizzhveIAMAAAAKGmQAAABAQYMMAAAAKGiQAQAAAAWb9OA3NunBDTbpwQU2W8EN5k2QmBv3bAeFvPbaa1qckpIiaqLk8BA26QEAAAC+0CADAAAAChpkAAAAQMEaZPiNNchwgzXIcIG1pHCDeQM3WIMMAAAA+EKDDAAAAChokAEAAAAFDTIAAACgaBNg/QXHcc6GYiCIeD1v4VrmTexyO2+YM7GNeQM3mDdwwzpvAvqKBQAAANDSscQCAAAAUNAgAwAAAAoaZAAAAEBBgwwAAAAoaJABAAAABQ0yAAAAoKBBBgAAABQ0yAAAAICCBhkAAABQ/Ae4t3NtxCmZ9wAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#autoencoder\n",
        "#encoder\n",
        "input_img = Input(shape=(28, 28, 1))  # adapt this if using `channels_first` image data format\n",
        "\n",
        "x = Conv2D(32, (3, 3), activation='relu', padding='same')(input_img)\n",
        "x = MaxPool2D((2, 2), padding='same')(x)\n",
        "x = Conv2D(32, (3, 3), activation='relu', padding='same')(x)\n",
        "encoded = MaxPool2D((2, 2), padding='same')(x)\n",
        "# at this point the representation is (7, 7, 32)\n",
        "#decoder..\n",
        "x = Conv2D(32, (3, 3), activation='relu', padding='same')(encoded)\n",
        "x = UpSampling2D((2, 2))(x)\n",
        "x = Conv2D(32, (3, 3), activation='relu', padding='same')(x)\n",
        "x = UpSampling2D((2, 2))(x)\n",
        "decoded = Conv2D(1, (3, 3), activation='sigmoid', padding='same')(x)"
      ],
      "metadata": {
        "id": "3AVTH-qcpxE6"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "autoencoder = Model(input_img, decoded)\n",
        "\n",
        "autoencoder.compile(optimizer='adadelta', loss='binary_crossentropy')"
      ],
      "metadata": {
        "id": "dEFwrEC5qn6N"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# train the convolutional autoencoder\n",
        "H = autoencoder.fit(\n",
        "\ttrainXNoisy, trainX,\n",
        "\tvalidation_data=(testXNoisy, testX),\n",
        "\tepochs=20,\n",
        "\tbatch_size=32)\n",
        "N = np.arange(0, 20)\n",
        "plt.style.use(\"ggplot\")\n",
        "plt.figure()\n",
        "plt.plot(N, H.history[\"loss\"], label=\"train_loss\")\n",
        "plt.plot(N, H.history[\"val_loss\"], label=\"val_loss\")\n",
        "plt.title(\"Training Loss and Accuracy\")\n",
        "plt.xlabel(\"Epoch #\")\n",
        "plt.ylabel(\"Loss/Accuracy\")\n",
        "plt.legend(loc=\"lower left\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 993
        },
        "id": "e5NNByF4qsae",
        "outputId": "b5310875-5d76-45fd-b248-23bafd22de6c"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/20\n",
            "1875/1875 [==============================] - 167s 88ms/step - loss: 0.5466 - val_loss: 0.4822\n",
            "Epoch 2/20\n",
            "1875/1875 [==============================] - 160s 85ms/step - loss: 0.4614 - val_loss: 0.4465\n",
            "Epoch 3/20\n",
            "1875/1875 [==============================] - 164s 87ms/step - loss: 0.4246 - val_loss: 0.4059\n",
            "Epoch 4/20\n",
            "1875/1875 [==============================] - 166s 89ms/step - loss: 0.3728 - val_loss: 0.3405\n",
            "Epoch 5/20\n",
            "1875/1875 [==============================] - 163s 87ms/step - loss: 0.3086 - val_loss: 0.2863\n",
            "Epoch 6/20\n",
            "1875/1875 [==============================] - 164s 87ms/step - loss: 0.2721 - val_loss: 0.2623\n",
            "Epoch 7/20\n",
            "1875/1875 [==============================] - 165s 88ms/step - loss: 0.2546 - val_loss: 0.2484\n",
            "Epoch 8/20\n",
            "1875/1875 [==============================] - 166s 89ms/step - loss: 0.2430 - val_loss: 0.2380\n",
            "Epoch 9/20\n",
            "1875/1875 [==============================] - 166s 89ms/step - loss: 0.2338 - val_loss: 0.2294\n",
            "Epoch 10/20\n",
            "1875/1875 [==============================] - 164s 88ms/step - loss: 0.2256 - val_loss: 0.2216\n",
            "Epoch 11/20\n",
            "1875/1875 [==============================] - 174s 93ms/step - loss: 0.2188 - val_loss: 0.2153\n",
            "Epoch 12/20\n",
            "1875/1875 [==============================] - 166s 89ms/step - loss: 0.2130 - val_loss: 0.2098\n",
            "Epoch 13/20\n",
            "1875/1875 [==============================] - 169s 90ms/step - loss: 0.2079 - val_loss: 0.2049\n",
            "Epoch 14/20\n",
            "1875/1875 [==============================] - 167s 89ms/step - loss: 0.2033 - val_loss: 0.2005\n",
            "Epoch 15/20\n",
            "1875/1875 [==============================] - 168s 89ms/step - loss: 0.1990 - val_loss: 0.1963\n",
            "Epoch 16/20\n",
            "1875/1875 [==============================] - 167s 89ms/step - loss: 0.1951 - val_loss: 0.1927\n",
            "Epoch 17/20\n",
            "1875/1875 [==============================] - 169s 90ms/step - loss: 0.1918 - val_loss: 0.1895\n",
            "Epoch 18/20\n",
            "1875/1875 [==============================] - 170s 91ms/step - loss: 0.1889 - val_loss: 0.1867\n",
            "Epoch 19/20\n",
            "1875/1875 [==============================] - 168s 90ms/step - loss: 0.1863 - val_loss: 0.1843\n",
            "Epoch 20/20\n",
            "1875/1875 [==============================] - 167s 89ms/step - loss: 0.1841 - val_loss: 0.1820\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# predict the reconstructed images for the original images\n",
        "pred = autoencoder.predict(testXNoisy)"
      ],
      "metadata": {
        "id": "d7q1tYUerCwY"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "## Visualizing our results\n",
        "\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(5):\n",
        "    plt.subplot(1, 5, i+1)\n",
        "    plt.xticks([]) # to remove x-axis  the [] empty list indicates this\n",
        "    plt.yticks([]) # to remove y-axis\n",
        "    plt.grid(False) # to remove grid\n",
        "    plt.imshow(testX[i].reshape(28, 28), cmap='gray') #display the image \n",
        "plt.tight_layout() # to have a proper space in the subplots\n",
        "plt.show()\n",
        "\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(5):\n",
        "    plt.subplot(1, 5, i+1)\n",
        "    plt.xticks([]) # to remove x-axis  the [] empty list indicates this\n",
        "    plt.yticks([]) # to remove y-axis\n",
        "    plt.grid(False) # to remove grid\n",
        "    plt.imshow(testXNoisy[i].reshape(28, 28), cmap='gray') #display the image \n",
        "plt.tight_layout() # to have a proper space in the subplots\n",
        "plt.show()\n",
        "\n",
        "# to visualise reconstructed images(output of autoencoder)\n",
        "plt.figure(figsize=(10,10))\n",
        "for i in range(5):\n",
        "    plt.subplot(1, 5, i+1)\n",
        "    plt.xticks([]) # to remove x-axis  the [] empty list indicates this\n",
        "    plt.yticks([]) # to remove y-axis\n",
        "    plt.grid(False) # to remove grid\n",
        "    plt.imshow(pred[i].reshape(28, 28), cmap='gray') #display the image \n",
        "plt.tight_layout() # to have a proper space in the subplots\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 452
        },
        "id": "MqcgIYYN4QnM",
        "outputId": "b2efbda3-0a93-4e3d-ffc5-0ac4af8b1e27"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 720x720 with 5 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "u6xiGQ7g4Xn9"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}