3.3 Guided Backpropagation.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3.3 Guided Backpropagation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Tensorflow Walkthrough"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1. Import Dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "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": [
    "import os\n",
    "\n",
    "from tensorflow.examples.tutorials.mnist import input_data\n",
    "from tensorflow.python.ops import nn_ops, gen_nn_ops\n",
    "from tensorflow.python.framework import ops\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "\n",
    "from models.models_3_3 import MNIST_CNN\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)\n",
    "\n",
    "images = mnist.train.images\n",
    "labels = mnist.train.labels\n",
    "\n",
    "logdir = './tf_logs/3_3_GBP/'\n",
    "ckptdir = logdir + 'model'\n",
    "\n",
    "if not os.path.exists(logdir):\n",
    "    os.mkdir(logdir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](./assets/3_3_GBP/fig1.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature:\n",
      " [[-2  2]\n",
      " [ 2 -2]]\n",
      "Gradient:\n",
      " [[-1 -1]\n",
      " [ 1  1]]\n",
      "Forward Pass:\n",
      " [[ 0.  2.]\n",
      " [ 2.  0.]]\n",
      "Backpropagation:\n",
      " [[ 0. -1.]\n",
      " [ 1.  0.]]\n",
      "Deconvolution:\n",
      " [[ 0.  0.]\n",
      " [ 1.  1.]]\n",
      "Guided Backpropagation:\n",
      " [[ 0.  0.]\n",
      " [ 1.  0.]]\n"
     ]
    }
   ],
   "source": [
    "# https://stackoverflow.com/questions/38340791/guided-back-propagation-in-tensorflow\n",
    "\n",
    "grad = tf.placeholder(tf.float32, [2,2])\n",
    "feat = tf.placeholder(tf.float32, [2,2])\n",
    "\n",
    "# Forward pass\n",
    "frwd = tf.nn.relu(feat)\n",
    "\n",
    "# Gradient calculation using backpropagation\n",
    "res1 = gen_nn_ops._relu_grad(grad, feat)\n",
    "\n",
    "# Gradient calculation using deconvolution\n",
    "res2 = tf.nn.relu(grad)\n",
    "\n",
    "# Gradient calculation using guided backpropagation\n",
    "res3 = tf.where(0. < grad, gen_nn_ops._relu_grad(grad, feat), tf.zeros(grad.get_shape()))\n",
    "\n",
    "sess = tf.InteractiveSession()\n",
    "\n",
    "grad_inp = np.array([[-1, -1], [1, 1]])\n",
    "feat_inp = np.array([[-2, 2], [2, -2]])\n",
    "\n",
    "f, r1, r2, r3 = sess.run([frwd, res1, res2, res3], feed_dict={grad: grad_inp, feat: feat_inp})\n",
    "\n",
    "print('Feature:\\n', feat_inp)\n",
    "print('Gradient:\\n', grad_inp)\n",
    "print('Forward Pass:\\n', f)\n",
    "print('Backpropagation:\\n', r1)\n",
    "print('Deconvolution:\\n', r2)\n",
    "print('Guided Backpropagation:\\n', r3)\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2. Building Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "with tf.name_scope('Classifier'):\n",
    "        \n",
    "    # Initialize neural network\n",
    "    DNN = MNIST_CNN('CNN')\n",
    "\n",
    "    # Setup training process\n",
    "    X = tf.placeholder(tf.float32, [None, 784], name='X')\n",
    "    Y = tf.placeholder(tf.float32, [None, 10], name='Y')\n",
    "\n",
    "    activations, logits = DNN(X)\n",
    "    \n",
    "    tf.add_to_collection('BP', X)\n",
    "    tf.add_to_collection('BP', logits)\n",
    "    \n",
    "    for activation in activations:\n",
    "        tf.add_to_collection('BP', activation)\n",
    "\n",
    "    cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=Y))\n",
    "\n",
    "    optimizer = tf.train.AdamOptimizer().minimize(cost, var_list=DNN.vars)\n",
    "\n",
    "    correct_prediction = tf.equal(tf.argmax(logits, 1), tf.argmax(Y, 1))\n",
    "    accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))\n",
    "\n",
    "cost_summary = tf.summary.scalar('Cost', cost)\n",
    "accuray_summary = tf.summary.scalar('Accuracy', accuracy)\n",
    "summary = tf.summary.merge_all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3. Training Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 0001 cost = 0.578529830 accuracy = 0.818581820\n",
      "Epoch: 0002 cost = 0.139815294 accuracy = 0.960454551\n",
      "Epoch: 0003 cost = 0.100711791 accuracy = 0.969490916\n",
      "Epoch: 0004 cost = 0.076209596 accuracy = 0.976836373\n",
      "Epoch: 0005 cost = 0.061706270 accuracy = 0.981472737\n",
      "Epoch: 0006 cost = 0.053644675 accuracy = 0.983345464\n",
      "Epoch: 0007 cost = 0.043971461 accuracy = 0.986981827\n",
      "Epoch: 0008 cost = 0.035710182 accuracy = 0.988854555\n",
      "Epoch: 0009 cost = 0.032316597 accuracy = 0.990181826\n",
      "Epoch: 0010 cost = 0.028523090 accuracy = 0.990963644\n",
      "Epoch: 0011 cost = 0.024217520 accuracy = 0.992163643\n",
      "Epoch: 0012 cost = 0.021784965 accuracy = 0.993127279\n",
      "Epoch: 0013 cost = 0.018662456 accuracy = 0.994200005\n",
      "Epoch: 0014 cost = 0.017762397 accuracy = 0.994490914\n",
      "Epoch: 0015 cost = 0.014130737 accuracy = 0.995454550\n",
      "Accuracy: 0.9865\n"
     ]
    }
   ],
   "source": [
    "sess = tf.InteractiveSession()\n",
    "sess.run(tf.global_variables_initializer())\n",
    "\n",
    "saver = tf.train.Saver()\n",
    "file_writer = tf.summary.FileWriter(logdir, tf.get_default_graph())\n",
    "\n",
    "# Hyper parameters\n",
    "training_epochs = 15\n",
    "batch_size = 100\n",
    "\n",
    "for epoch in range(training_epochs):\n",
    "    total_batch = int(mnist.train.num_examples / batch_size)\n",
    "    avg_cost = 0\n",
    "    avg_acc = 0\n",
    "    \n",
    "    for i in range(total_batch):\n",
    "        batch_xs, batch_ys = mnist.train.next_batch(batch_size)\n",
    "        _, c, a, summary_str = sess.run([optimizer, cost, accuracy, summary], feed_dict={X: batch_xs, Y: batch_ys})\n",
    "        avg_cost += c / total_batch\n",
    "        avg_acc += a / total_batch\n",
    "        \n",
    "        file_writer.add_summary(summary_str, epoch * total_batch + i)\n",
    "\n",
    "    print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.9f}'.format(avg_cost), 'accuracy =', '{:.9f}'.format(avg_acc))\n",
    "    \n",
    "    saver.save(sess, ckptdir)\n",
    "\n",
    "print('Accuracy:', sess.run(accuracy, feed_dict={X: mnist.test.images, Y: mnist.test.labels}))\n",
    "\n",
    "sess.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 4. Restoring Graph"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./tf_logs/2_5_GBP/model\n"
     ]
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "\n",
    "@ops.RegisterGradient(\"GuidedRelu\")\n",
    "def _GuidedReluGrad(op, grad):\n",
    "    return tf.where(0. < grad, gen_nn_ops._relu_grad(grad, op.outputs[0]), tf.zeros(tf.shape(grad)))\n",
    "\n",
    "sess = tf.InteractiveSession()\n",
    "\n",
    "g = tf.get_default_graph()\n",
    "with g.gradient_override_map({'Relu': 'GuidedRelu'}):\n",
    "    new_saver = tf.train.import_meta_graph(ckptdir + '.meta')\n",
    "\n",
    "new_saver.restore(sess, tf.train.latest_checkpoint(logdir))\n",
    "\n",
    "activations = tf.get_collection('BP')\n",
    "X = activations[0]\n",
    "logits = activations[1]\n",
    "\n",
    "hmaps = [tf.gradients(logits[:,i,None], X)[0] for i in range(10)]\n",
    "\n",
    "sample_imgs = [images[np.argmax(labels, axis=1) == i][5] for i in range(10)]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 5. Displaying Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAtUAAAQwCAYAAAAq6qJgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nqykAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f0b16fccc0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "features = [sess.run(hmaps[i], feed_dict={X: sample_imgs[i][None]}) for i in range(10)]\n",
    "\n",
    "plt.figure(figsize=(15,15))\n",
    "for i in range(5):\n",
    "    plt.subplot(5, 2, 2 * i + 1)\n",
    "    plt.imshow(np.reshape(features[2 * i], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i))\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.colorbar()\n",
    "    \n",
    "    plt.subplot(5, 2, 2 * i + 2)\n",
    "    plt.imshow(np.reshape(features[2 * i + 1], [28, 28]), cmap='gray', interpolation='none')\n",
    "    plt.title('Digit: {}'.format(2 * i + 1))\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.colorbar()\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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"
  },
  "latex_envs": {
   "LaTeX_envs_menu_present": true,
   "autoclose": false,
   "autocomplete": true,
   "bibliofile": "biblio.bib",
   "cite_by": "apalike",
   "current_citInitial": 1,
   "eqLabelWithNumbers": true,
   "eqNumInitial": 1,
   "hotkeys": {
    "equation": "Ctrl-E",
    "itemize": "Ctrl-I"
   },
   "labels_anchors": false,
   "latex_user_defs": false,
   "report_style_numbering": false,
   "user_envs_cfg": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}