Project_6_Image_reconstruction_using_Autoencoder.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyNsevpSBWVh1jL9lODyLviR",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Aaina26/Deep-Learning-Unschool/blob/main/Project_6_Image_reconstruction_using_Autoencoder.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Vanilla Autoencoder**"
      ],
      "metadata": {
        "id": "efM4_xBagh42"
      }
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "Vsn42cuBIqoh"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.layers import Dense, Input, Flatten, Reshape, LeakyReLU as LR, Activation, Dropout\n",
        "from tensorflow.keras.models import Model, Sequential\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "from IPython import display"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#loading data\n",
        "fashion_mnist=tf.keras.datasets.fashion_mnist\n",
        "(x_train,y_train),(x_test,y_test)=fashion_mnist.load_data()\n",
        "x_train,x_test=x_train/255.0,x_test/255.0 #normalize data\n",
        "print(\"x_train.shape\",x_train.shape)\n",
        "print(\"y_tarin.shape\",y_train.shape)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "HakdmztXKRVe",
        "outputId": "4a07d10c-d77a-47a8-d769-a0eb3edea587"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
            "29515/29515 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
            "26421880/26421880 [==============================] - 1s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
            "5148/5148 [==============================] - 0s 0us/step\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
            "4422102/4422102 [==============================] - 0s 0us/step\n",
            "x_train.shape (60000, 28, 28)\n",
            "y_tarin.shape (60000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "LATENT_SIZE=31"
      ],
      "metadata": {
        "id": "gNsVm8trL4cq"
      },
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#build encoder\n",
        "encoder=Sequential([\n",
        "    Flatten(input_shape=(28,28)),\n",
        "    Dense(512),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(256),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(64),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(LATENT_SIZE),\n",
        "    LR()\n",
        "])"
      ],
      "metadata": {
        "id": "7Xn1SttqMRk6"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#DECODER\n",
        "decoder=Sequential([\n",
        "    Dense(64,input_shape=(LATENT_SIZE,)),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(128),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(256),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(512),\n",
        "    LR(),\n",
        "    Dropout(0.5),\n",
        "    Dense(784), #TO GET 28*28 FEATURES\n",
        "    Activation(\"sigmoid\"),\n",
        "    Reshape((28,28))\n",
        "])"
      ],
      "metadata": {
        "id": "k-KKOudjNcqo"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "img=Input(shape=(28,28))\n"
      ],
      "metadata": {
        "id": "DyX-PAKjOraQ"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "latent_vector=encoder(img)\n",
        "output=decoder(latent_vector)\n"
      ],
      "metadata": {
        "id": "4A-osfjVPSnb"
      },
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model=Model(inputs=img,outputs=output)\n",
        "model.compile(\"nadam\",loss=\"binary_crossentropy\")"
      ],
      "metadata": {
        "id": "dZjgznPHPfAQ"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "autoencoder_train=model.fit(x_train, x_train, epochs=60)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fBjgbqqbP7-i",
        "outputId": "11f25a05-1338-4833-aa19-abc12e33a32f"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/60\n",
            "1875/1875 [==============================] - 18s 5ms/step - loss: 0.3909\n",
            "Epoch 2/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3619\n",
            "Epoch 3/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3554\n",
            "Epoch 4/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3514\n",
            "Epoch 5/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3492\n",
            "Epoch 6/60\n",
            "1875/1875 [==============================] - 9s 5ms/step - loss: 0.3477\n",
            "Epoch 7/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3468\n",
            "Epoch 8/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3461\n",
            "Epoch 9/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3454\n",
            "Epoch 10/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3452\n",
            "Epoch 11/60\n",
            "1875/1875 [==============================] - 9s 5ms/step - loss: 0.3448\n",
            "Epoch 12/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3447\n",
            "Epoch 13/60\n",
            "1875/1875 [==============================] - 12s 6ms/step - loss: 0.3444\n",
            "Epoch 14/60\n",
            "1875/1875 [==============================] - 13s 7ms/step - loss: 0.3441\n",
            "Epoch 15/60\n",
            "1875/1875 [==============================] - 14s 8ms/step - loss: 0.3439\n",
            "Epoch 16/60\n",
            "1875/1875 [==============================] - 12s 7ms/step - loss: 0.3438\n",
            "Epoch 17/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3441\n",
            "Epoch 18/60\n",
            "1875/1875 [==============================] - 11s 6ms/step - loss: 0.3437\n",
            "Epoch 19/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3436\n",
            "Epoch 20/60\n",
            "1875/1875 [==============================] - 11s 6ms/step - loss: 0.3435\n",
            "Epoch 21/60\n",
            "1875/1875 [==============================] - 13s 7ms/step - loss: 0.3434\n",
            "Epoch 22/60\n",
            "1875/1875 [==============================] - 15s 8ms/step - loss: 0.3432\n",
            "Epoch 23/60\n",
            "1875/1875 [==============================] - 12s 7ms/step - loss: 0.3433\n",
            "Epoch 24/60\n",
            "1875/1875 [==============================] - 11s 6ms/step - loss: 0.3433\n",
            "Epoch 25/60\n",
            "1875/1875 [==============================] - 10s 6ms/step - loss: 0.3431\n",
            "Epoch 26/60\n",
            "1875/1875 [==============================] - 12s 6ms/step - loss: 0.3432\n",
            "Epoch 27/60\n",
            "1875/1875 [==============================] - 13s 7ms/step - loss: 0.3430\n",
            "Epoch 28/60\n",
            "1875/1875 [==============================] - 13s 7ms/step - loss: 0.3433\n",
            "Epoch 29/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3430\n",
            "Epoch 30/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3431\n",
            "Epoch 31/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3430\n",
            "Epoch 32/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3430\n",
            "Epoch 33/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3429\n",
            "Epoch 34/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3429\n",
            "Epoch 35/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3428\n",
            "Epoch 36/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3429\n",
            "Epoch 37/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3428\n",
            "Epoch 38/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3427\n",
            "Epoch 39/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3426\n",
            "Epoch 40/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3426\n",
            "Epoch 41/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3426\n",
            "Epoch 42/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3422\n",
            "Epoch 43/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3427\n",
            "Epoch 44/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 45/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3424\n",
            "Epoch 46/60\n",
            "1875/1875 [==============================] - 10s 6ms/step - loss: 0.3424\n",
            "Epoch 47/60\n",
            "1875/1875 [==============================] - 11s 6ms/step - loss: 0.3424\n",
            "Epoch 48/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3426\n",
            "Epoch 49/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 50/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3422\n",
            "Epoch 51/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 52/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 53/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3426\n",
            "Epoch 54/60\n",
            "1875/1875 [==============================] - 10s 6ms/step - loss: 0.3424\n",
            "Epoch 55/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 56/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3421\n",
            "Epoch 57/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3423\n",
            "Epoch 58/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3424\n",
            "Epoch 59/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3425\n",
            "Epoch 60/60\n",
            "1875/1875 [==============================] - 10s 5ms/step - loss: 0.3421\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#plot randomly predicted images\n",
        "#more number of epochs, less blurring\n",
        "fig, axs=plt.subplots(5,5)\n",
        "rand=x_test[np.random.randint(0, 10000,25)].reshape((5,5,1,28,28))\n",
        "for i in range(5):\n",
        "  for j in range(5):\n",
        "    axs[i,j].imshow(model.predict(rand[i,j])[0],cmap='gray')\n",
        "    axs[i,j].axis(\"off\")\n",
        "\n",
        "plt.subplots_adjust(wspace=0,hspace=0)\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 682
        },
        "id": "vRvulsJsSiCI",
        "outputId": "129dd239-65f1-44aa-a1ef-c28ec66d2b03"
      },
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 20ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 30ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 21ms/step\n",
            "1/1 [==============================] - 0s 24ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n",
            "1/1 [==============================] - 0s 22ms/step\n",
            "1/1 [==============================] - 0s 23ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 25 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loss=autoencoder_train.history['loss']\n",
        "epochs=range(60)\n",
        "plt.figure()\n",
        "plt.plot(epochs,loss,'bo',label='Training loss')\n",
        "plt.title('Training loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "Q3ckDS-jUDsS",
        "outputId": "4ede3a84-0b69-495e-84e3-3204624dc5ee"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "**Convolutional Autoencoder**"
      ],
      "metadata": {
        "id": "7LT1olnrgMju"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "batch_size=128\n",
        "epochs=50\n",
        "inChannel=1\n",
        "x, y=28,28\n",
        "input_img=Input(shape=(x,y,inChannel))\n"
      ],
      "metadata": {
        "id": "tLmjcAG0gu-N"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "train_data=x_train\n",
        "train_x,valid_x,train_ground,valid_ground=train_test_split(train_data,train_data,test_size=0.2,random_state=0)"
      ],
      "metadata": {
        "id": "oMfcqMkShMOZ"
      },
      "execution_count": 16,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.layers import Conv2D,MaxPooling2D,UpSampling2D\n",
        "def autoencoder(input_img):\n",
        "  #encoder\n",
        "  conv1=Conv2D(32,(3,3),activation='relu',padding='same')(input_img)\n",
        "  pool1=MaxPooling2D(pool_size=(2,2))(conv1)\n",
        "  conv2=Conv2D(64,(3,3),activation='relu',padding='same')(pool1)\n",
        "  pool2=MaxPooling2D(pool_size=(2,2))(conv2)\n",
        "  conv3=Conv2D(128,(3,3),activation='relu',padding='same')(pool2)\n",
        "  #decoder\n",
        "  conv4=Conv2D(128,(3,3),activation='relu',padding='same')(conv3)\n",
        "  up1=UpSampling2D((2,2))(conv4)\n",
        "  conv5=Conv2D(64,(3,3),activation='relu',padding='same')(up1)\n",
        "  up2=UpSampling2D((2,2))(conv5)\n",
        "  decoded=Conv2D(1,(3,3),activation='sigmoid',padding='same')(up2)\n",
        "  return decoded\n"
      ],
      "metadata": {
        "id": "1ov15COWjgYV"
      },
      "execution_count": 17,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from keras.optimizers import RMSprop\n",
        "autoencoder=Model(input_img,autoencoder(input_img))\n",
        "autoencoder.compile(loss='mean_squared_error',optimizer=RMSprop())"
      ],
      "metadata": {
        "id": "Ko1PadK2mHxH"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "autoencoder_train=autoencoder.fit(train_x,train_ground,batch_size=batch_size,epochs=epochs,verbose=1,validation_data=(valid_x,valid_ground))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dc72k5Ovm0w7",
        "outputId": "03776da0-86a9-44d1-a957-6b6cc6a06177"
      },
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "375/375 [==============================] - 11s 13ms/step - loss: 0.0230 - val_loss: 0.0118\n",
            "Epoch 2/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0095 - val_loss: 0.0072\n",
            "Epoch 3/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0072 - val_loss: 0.0061\n",
            "Epoch 4/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0059 - val_loss: 0.0048\n",
            "Epoch 5/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0051 - val_loss: 0.0048\n",
            "Epoch 6/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0046 - val_loss: 0.0040\n",
            "Epoch 7/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0042 - val_loss: 0.0043\n",
            "Epoch 8/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0039 - val_loss: 0.0036\n",
            "Epoch 9/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0037 - val_loss: 0.0037\n",
            "Epoch 10/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0035 - val_loss: 0.0034\n",
            "Epoch 11/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0033 - val_loss: 0.0032\n",
            "Epoch 12/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0032 - val_loss: 0.0028\n",
            "Epoch 13/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0031 - val_loss: 0.0033\n",
            "Epoch 14/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0030 - val_loss: 0.0029\n",
            "Epoch 15/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0029 - val_loss: 0.0031\n",
            "Epoch 16/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0029 - val_loss: 0.0028\n",
            "Epoch 17/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0028 - val_loss: 0.0031\n",
            "Epoch 18/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0027 - val_loss: 0.0028\n",
            "Epoch 19/50\n",
            "375/375 [==============================] - 6s 16ms/step - loss: 0.0027 - val_loss: 0.0030\n",
            "Epoch 20/50\n",
            "375/375 [==============================] - 5s 15ms/step - loss: 0.0026 - val_loss: 0.0028\n",
            "Epoch 21/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0026 - val_loss: 0.0024\n",
            "Epoch 22/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0026 - val_loss: 0.0023\n",
            "Epoch 23/50\n",
            "375/375 [==============================] - 5s 14ms/step - loss: 0.0025 - val_loss: 0.0028\n",
            "Epoch 24/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0025 - val_loss: 0.0024\n",
            "Epoch 25/50\n",
            "375/375 [==============================] - 5s 14ms/step - loss: 0.0024 - val_loss: 0.0025\n",
            "Epoch 26/50\n",
            "375/375 [==============================] - 6s 16ms/step - loss: 0.0024 - val_loss: 0.0024\n",
            "Epoch 27/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0024 - val_loss: 0.0022\n",
            "Epoch 28/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0024 - val_loss: 0.0023\n",
            "Epoch 29/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0023 - val_loss: 0.0024\n",
            "Epoch 30/50\n",
            "375/375 [==============================] - 5s 14ms/step - loss: 0.0023 - val_loss: 0.0027\n",
            "Epoch 31/50\n",
            "375/375 [==============================] - 5s 14ms/step - loss: 0.0023 - val_loss: 0.0023\n",
            "Epoch 32/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0023 - val_loss: 0.0023\n",
            "Epoch 33/50\n",
            "375/375 [==============================] - 6s 15ms/step - loss: 0.0022 - val_loss: 0.0023\n",
            "Epoch 34/50\n",
            "375/375 [==============================] - 6s 16ms/step - loss: 0.0022 - val_loss: 0.0021\n",
            "Epoch 35/50\n",
            "375/375 [==============================] - 6s 15ms/step - loss: 0.0022 - val_loss: 0.0022\n",
            "Epoch 36/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0022 - val_loss: 0.0023\n",
            "Epoch 37/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0022 - val_loss: 0.0024\n",
            "Epoch 38/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0021 - val_loss: 0.0024\n",
            "Epoch 39/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0021 - val_loss: 0.0021\n",
            "Epoch 40/50\n",
            "375/375 [==============================] - 4s 12ms/step - loss: 0.0021 - val_loss: 0.0021\n",
            "Epoch 41/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0021 - val_loss: 0.0023\n",
            "Epoch 42/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0021 - val_loss: 0.0022\n",
            "Epoch 43/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0021 - val_loss: 0.0021\n",
            "Epoch 44/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0021 - val_loss: 0.0019\n",
            "Epoch 45/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0020 - val_loss: 0.0020\n",
            "Epoch 46/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0020 - val_loss: 0.0022\n",
            "Epoch 47/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0020 - val_loss: 0.0020\n",
            "Epoch 48/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0020 - val_loss: 0.0020\n",
            "Epoch 49/50\n",
            "375/375 [==============================] - 5s 13ms/step - loss: 0.0020 - val_loss: 0.0022\n",
            "Epoch 50/50\n",
            "375/375 [==============================] - 5s 12ms/step - loss: 0.0020 - val_loss: 0.0021\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loss=autoencoder_train.history['loss']\n",
        "val_loss=autoencoder_train.history['val_loss']\n",
        "epochs=range(50)\n",
        "plt.figure()\n",
        "plt.plot(epochs,loss,'bo',label='Training loss')\n",
        "plt.plot(epochs,val_loss,'b',label='Validation loss',color='red')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "id": "986WGWq_nf2a",
        "outputId": "0f2f8b66-4f2f-42e3-8fbb-97d4206c507c"
      },
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-24-7fe7b74e75a4>:6: UserWarning: color is redundantly defined by the 'color' keyword argument and the fmt string \"b\" (-> color=(0.0, 0.0, 1.0, 1)). The keyword argument will take precedence.\n",
            "  plt.plot(epochs,val_loss,'b',label='Validation loss',color='red')\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#predict the images\n",
        "pred=autoencoder.predict(x_test)\n",
        "plt.figure(figsize=(20,4))\n",
        "print(\"Reconstruction of Test Images\")\n",
        "for i in range(10,20,1):\n",
        "  plt.subplot(2,10,i+1)\n",
        "  plt.imshow(pred[i, ..., 0],cmap='gray')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180
        },
        "id": "M0V1FagKocVt",
        "outputId": "7bdab536-f24c-4a98-de67-755e27704d0a"
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 3ms/step\n",
            "Reconstruction of Test Images\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1440x288 with 10 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}