Denoising-Autoencoder/tf_dae_mnist_reconstruction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# De-noising Feedforward Autoencoder\n",
    "##### MNIST reconstruction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Util functions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "function to plot the images after during testing phase"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def plot_images(images, title, no_i_x, no_i_y=3):\n",
    "    fig = plt.figure(figsize=(10, 15))\n",
    "    fig.canvas.set_window_title(title)\n",
    "    images = np.array(images).reshape(-1, 28, 28)\n",
    "    for i in range(no_i_x):\n",
    "        for j in range(no_i_y):\n",
    "            ax = fig.add_subplot(no_i_x, no_i_y, no_i_x * j + (i + 1))\n",
    "            ax.matshow(images[no_i_x * j + i], cmap=\"gray\")\n",
    "            plt.xticks(np.array([]))\n",
    "            plt.yticks(np.array([]))\n",
    "\n",
    "            if j == 0 and i == 0:\n",
    "                ax.set_title(\"Real\")\n",
    "            elif j == 0 and i == 1:\n",
    "                ax.set_title(\"Distorted\")\n",
    "            elif j == 0 and i == 2:\n",
    "                ax.set_title(\"Reconstructed\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "function to add noise to input data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def add_noise(data, mean=0, stddev=0.2):\n",
    "    try:\n",
    "        noise = np.random.normal(mean, stddev, data.shape)\n",
    "    except ValueError:\n",
    "        noise = np.zeros_like(data)\n",
    "\n",
    "    noisy_data = data + noise\n",
    "    clipped_noisy_data = np.clip(noisy_data, 0.0, 1.0)\n",
    "\n",
    "    return clipped_noisy_data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "load the mmist dataset from tensorflow.examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Extracting MNIST_data/train-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/train-labels-idx1-ubyte.gz\n",
      "Extracting MNIST_data/t10k-images-idx3-ubyte.gz\n",
      "Extracting MNIST_data/t10k-labels-idx1-ubyte.gz\n"
     ]
    }
   ],
   "source": [
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "features_train, features_test, features_valid = \\\n",
    "    mnist.train.images, mnist.test.images, mnist.validation.images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Neural Network Model\n",
    "2 Encoder layers + 2 Decoder layers (all dense neural layers)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<br>Hyper-parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_input_layer = features_train.shape[1]\n",
    "n_enc_hidden_1 = 400\n",
    "n_enc_hidden_2 = 100  # 7.84 times compression\n",
    "n_dec_hidden_1 = 100\n",
    "n_dec_hidden_2 = 400\n",
    "n_output_layer = features_train.shape[1]\n",
    "\n",
    "learning_rate = 0.001\n",
    "\n",
    "n_epoch = 10\n",
    "batch_size = 100\n",
    "\n",
    "test_disp = 10  # no of images in plot"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "input/output placeholders"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X = tf.placeholder(tf.float32, [None, n_input_layer])\n",
    "Y = tf.placeholder(tf.float32, [None, n_output_layer])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Weights and biases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "enc_layer_1 = {\n",
    "    'weights': tf.Variable(tf.random_normal([n_input_layer, n_enc_hidden_1], stddev=0.1)),\n",
    "    'biases': tf.Variable(tf.random_normal([n_enc_hidden_1], stddev=0.1))\n",
    "}\n",
    "enc_layer_2 = {\n",
    "    'weights': tf.Variable(tf.random_normal([n_enc_hidden_1, n_enc_hidden_2], stddev=0.1)),\n",
    "    'biases': tf.Variable(tf.random_normal([n_enc_hidden_2], stddev=0.1))\n",
    "}\n",
    "dec_layer_1 = {\n",
    "    'weights': tf.Variable(tf.random_normal([n_dec_hidden_1, n_dec_hidden_2], stddev=0.1)),\n",
    "    'biases': tf.Variable(tf.random_normal([n_dec_hidden_2], stddev=0.1))\n",
    "}\n",
    "dec_layer_2 = {\n",
    "    'weights': tf.Variable(tf.random_normal([n_dec_hidden_2, n_output_layer], stddev=0.1)),\n",
    "    'biases': tf.Variable(tf.random_normal([n_output_layer], stddev=0.1))\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Encoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "h_enc_1 = tf.add(tf.matmul(X, enc_layer_1['weights']), enc_layer_1['biases'])\n",
    "enc_1 = tf.nn.sigmoid(h_enc_1)\n",
    "\n",
    "h_enc_2 = tf.add(tf.matmul(enc_1, enc_layer_2['weights']), enc_layer_2['biases'])\n",
    "enc_2 = tf.nn.tanh(h_enc_2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Decoder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "h_dec_1 = tf.add(tf.matmul(enc_2, dec_layer_1['weights']), dec_layer_1['biases'])\n",
    "dec_1 = tf.nn.tanh(h_dec_1)\n",
    "\n",
    "h_dec_2 = tf.add(tf.matmul(dec_1, dec_layer_2['weights']), dec_layer_2['biases'])\n",
    "dec_2 = tf.nn.sigmoid(h_dec_2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Error and Optimizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# mean-squared error\n",
    "cost = tf.losses.mean_squared_error(labels=Y, predictions=dec_2)\n",
    "\n",
    "# adam-optimizer\n",
    "optimizer = tf.train.AdamOptimizer(learning_rate).minimize(cost)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Make Batches"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "n_batch = features_train.shape[0] // batch_size\n",
    "batched_data = np.split(features_train, n_batch)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Start session"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "*********** Train ***********\n",
      "Epoch: 0  Train-error: 0.0326655690457  Validation-error: 0.0158167\n",
      "Epoch: 1  Train-error: 0.012484143969  Validation-error: 0.0101491\n",
      "Epoch: 2  Train-error: 0.00908060977659  Validation-error: 0.00839202\n",
      "Epoch: 3  Train-error: 0.00768788457001  Validation-error: 0.00725421\n",
      "Epoch: 4  Train-error: 0.0068269499713  Validation-error: 0.00660782\n",
      "Epoch: 5  Train-error: 0.00626942924309  Validation-error: 0.00626707\n",
      "Epoch: 6  Train-error: 0.00589494798248  Validation-error: 0.00591861\n",
      "Epoch: 7  Train-error: 0.00564208484509  Validation-error: 0.00560433\n",
      "Epoch: 8  Train-error: 0.00545295548879  Validation-error: 0.00550049\n",
      "Epoch: 9  Train-error: 0.00528165180405  Validation-error: 0.00549375\n",
      "\n",
      "*********** Test ***********\n",
      "Test Error: 0.00539585\n"
     ]
    }
   ],
   "source": [
    "with tf.Session() as sess:\n",
    "    tf.global_variables_initializer().run()\n",
    "    \n",
    "    print(\"*********** Train ***********\")\n",
    "\n",
    "    # Epoch-training\n",
    "    for epoch in range(n_epoch):\n",
    "        tr_err = []\n",
    "\n",
    "        # Batch training\n",
    "        for b_idx in range(n_batch):\n",
    "            noisy_data = add_noise(batched_data[b_idx])\n",
    "            e, _ = sess.run([cost, optimizer], feed_dict={X: noisy_data, Y: batched_data[b_idx]})\n",
    "\n",
    "            tr_err.append(e)\n",
    "            \n",
    "        noisy_vaild_data = add_noise(features_valid)\n",
    "        val_err = cost.eval({X: noisy_vaild_data, Y: features_valid})\n",
    "\n",
    "        train_err = sum(tr_err) / len(tr_err)\n",
    "        print(\"Epoch:\", epoch, \" Train-error:\", train_err, \" Validation-error:\", val_err)\n",
    "        \n",
    "    print(\"\\n*********** Test ***********\")\n",
    "\n",
    "    # Test the model on test data and try to reconstruct it\n",
    "    noisy_test = add_noise(features_test)\n",
    "    reconstructed_imgs, err = sess.run([dec_2, cost], feed_dict={X: noisy_test, Y: features_test})\n",
    "    disp_imgs = []\n",
    "    for i in range(test_disp):\n",
    "        disp_imgs.append(features_test[i])\n",
    "        disp_imgs.append(noisy_test[i])\n",
    "        disp_imgs.append(reconstructed_imgs[i])\n",
    "\n",
    "    # Plot original, noisy and reconstructed images\n",
    "    print(\"Test Error: %.8f\" % err)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### plot output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAekAAANUCAYAAACXBdUVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFDtpY4wxpqLYSRtjjDEV5f8ApDA9tpCelwcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fbcea9a2cc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_images(disp_imgs, \"De-noising Auto-encoder\", test_disp)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}