awjuliani · January 28, 2021 10:18 · Dec 9, 2016 · Apr 6, 2016
diff --git a/Deep Layer Visualization.ipynb b/Deep Layer Visualization.ipynb
diff --git a/Deep Layer Visualization.ipynb b/Deep Layer Visualization.ipynb
@@ -0,0 +1,383 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Deep Layer Visualization Tutorial\n",
+    "\n",
+    "\n",
+    "First we import the needed libraries."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [],
+   "source": [
+    "import numpy as np \n",
+    "import matplotlib as mp\n",
+    "%matplotlib inline\n",
+    "import matplotlib.pyplot as plt\n",
+    "import tensorflow as tf"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Next we import the MNIST data files we are going to be classifying. This database contains images of thousands of handwritten digits, and their proper labels. For convenience I am using a script from Google, which can be download [here](https://github.com/tensorflow/tensorflow/blob/r0.7/tensorflow/examples/tutorials/mnist/input_data.py). Just add it to your working directory, and it will download the MNIST database for you. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {
+    "collapsed": false
+   },
+   "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 input_data\n",
+    "mnist = input_data.read_data_sets(\"MNIST_data/\", one_hot=True)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Next we define our convolutional network. It will be a network with three sets of convolution -> pooling layers, followed by a fully connected softmax layer. I have choosen 5,5,20 to begin with. Feel free to adjust the number of convolutional filters at each layer. It is these filters we will be visualizing, so we can see in realtime what features are learned from the dataset with more or less filters."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 15,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [],
+   "source": [
+    "sess = tf.Session()\n",
+    "\n",
+    "x = tf.placeholder(tf.float32, [None, 784],name=\"x-in\")\n",
+    "y_ = tf.placeholder(tf.float32, [None, 10],name=\"y-in\")\n",
+    "\n",
+    "def weight_variable(shape):\n",
+    "    initial = tf.truncated_normal(shape, stddev=0.1)\n",
+    "    return tf.Variable(initial)\n",
+    "\n",
+    "def bias_variable(shape):\n",
+    "    initial = tf.constant(0.1, shape=shape)\n",
+    "    return tf.Variable(initial)\n",
+    "\n",
+    "def conv2d(x, W):\n",
+    "    return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')\n",
+    "\n",
+    "def max_pool_2x2(x):\n",
+    "    return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],strides=[1, 2, 2, 1], padding='SAME')\n",
+    "\n",
+    "x_image = tf.reshape(x, [-1,28,28,1])\n",
+    "W_conv1 = weight_variable([5, 5, 1, 5])\n",
+    "b_conv1 = bias_variable([5])\n",
+    "h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)\n",
+    "h_pool1 = max_pool_2x2(h_conv1)\n",
+    "\n",
+    "W_conv2 = weight_variable([5, 5, 5, 5])\n",
+    "b_conv2 = bias_variable([5])\n",
+    "h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)\n",
+    "h_pool2 = max_pool_2x2(h_conv2)\n",
+    "\n",
+    "W_conv3 = weight_variable([5, 5, 5, 20])\n",
+    "b_conv3 = bias_variable([20])\n",
+    "h_conv3 = tf.nn.relu(conv2d(h_pool2, W_conv3) + b_conv3)\n",
+    "\n",
+    "W_fc1 = weight_variable([7 * 7 * 20, 10])\n",
+    "b_fc1 = bias_variable([10])\n",
+    "h_conv3_flat = tf.reshape(h_conv3, [-1, 7*7*20])\n",
+    "keep_prob = tf.placeholder(\"float\")\n",
+    "h_conv3_drop = tf.nn.dropout(h_conv3_flat, keep_prob)\n",
+    "y_conv = tf.nn.softmax(tf.matmul(h_conv3_drop, W_fc1) + b_fc1)\n",
+    "\n",
+    "cross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))\n",
+    "correct_prediction = tf.equal(tf.argmax(y_conv,1), tf.argmax(y_,1))\n",
+    "accuracy = tf.reduce_mean(tf.cast(correct_prediction, \"float\"))\n",
+    "train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "We then train the network using batch gradient descent with Adam optimization. Feel free to adjust the batch size and number of iterations to see how it effects the model accuracy."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {
+    "collapsed": false,
+    "scrolled": false
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "step 0, training accuracy 0.06\n",
+      "step 100, training accuracy 0.26\n",
+      "step 200, training accuracy 0.44\n",
+      "step 300, training accuracy 0.48\n",
+      "step 400, training accuracy 0.68\n",
+      "step 500, training accuracy 0.66\n",
+      "step 600, training accuracy 0.86\n",
+      "step 700, training accuracy 0.86\n",
+      "step 800, training accuracy 0.78\n",
+      "step 900, training accuracy 0.84\n",
+      "step 1000, training accuracy 0.9\n"
+     ]
+    }
+   ],
+   "source": [
+    "sess.run(tf.initialize_all_variables())\n",
+    "iterations = 0\n",
+    "batchSize = 50\n",
+    "while iterations < 1001:\n",
+    "    batch = mnist.train.next_batch(batchSize)\n",
+    "    train_step.run(session=sess, feed_dict={x:batch[0],y_:batch[1], keep_prob:0.5})\n",
+    "    if iterations%100 == 0:\n",
+    "        trainAccuracy = accuracy.eval(session=sess, feed_dict={x:batch[0],y_:batch[1], keep_prob:1.0})\n",
+    "        print(\"step %d, training accuracy %g\"%(iterations, trainAccuracy))\n",
+    "    iterations += 1"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "test accuracy 0.8869\n"
+     ]
+    }
+   ],
+   "source": [
+    "testAccuracy = accuracy.eval(session=sess, feed_dict={x:mnist.test.images,y_:mnist.test.labels, keep_prob:1.0})\n",
+    "print(\"test accuracy %g\"%(testAccuracy))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now we define a couple functions that will allow us to visualize the network. The first gets the activations at a given layer for a given input image. The second plots those activations in a grid."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [],
+   "source": [
+    "def getActivations(layer,stimuli):\n",
+    "    units = layer.eval(session=sess,feed_dict={x:np.reshape(stimuli,[1,784],order='F'),keep_prob:1.0})\n",
+    "    plotNNFilter(units)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {
+    "collapsed": false
+   },
+   "outputs": [],
+   "source": [
+    "def plotNNFilter(units):\n",
+    "    filters = units.shape[3]\n",
+    "    plt.figure(1, figsize=(20,20))\n",
+    "    for i in xrange(0,filters):\n",
+    "        plt.subplot(7,6,i+1)\n",
+    "        plt.title('Filter ' + str(i))\n",
+    "        plt.imshow(units[0,:,:,i], interpolation=\"nearest\", cmap=\"gray\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now we can choose an image to pass through the network to visualize the network activation, and look at the raw pixels of that image."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "metadata": {
+    "collapsed": false,
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "<matplotlib.image.AxesImage at 0x1143d7390>"
+      ]
+     },
+     "execution_count": 23,
+     "metadata": {},
+     "output_type": "execute_result"
+    },
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP4AAAD8CAYAAABXXhlaAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIvhAQgQfSIjgAwkRfCAhgg8kRPCBhP4HeOd/0sgDU6QAAAAASUVORK5CYII=\n",
+      "text/plain": [
+       "<matplotlib.figure.Figure at 0x1143e0190>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "imageToUse = mnist.test.images[0]\n",
+    "plt.imshow(np.reshape(imageToUse,[28,28]), interpolation=\"nearest\", cmap=\"gray\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Now we can look at how that image activates the neurons of the first convolutional layer. Notice how each filter has learned to activate optimally for different features of the image. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "metadata": {
+    "collapsed": false,
+    "scrolled": false
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6QAAACzCAYAAACehNFjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nrkJggg==\n",
+      "text/plain": [
+       "<matplotlib.figure.Figure at 0x115ca2390>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "getActivations(h_conv1,imageToUse)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "We can do this again for the second layer..."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "metadata": {
+    "collapsed": false,
+    "scrolled": true
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6QAAACzCAYAAACehNFjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmJACAAAAABLx/wF+G/BlfOakpAAAAABJRU5ErkJggg==\n",
+      "text/plain": [
+       "<matplotlib.figure.Figure at 0x114daa210>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "getActivations(h_conv2,imageToUse)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "...and the third."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "metadata": {
+    "collapsed": false,
+    "scrolled": false
+   },
+   "outputs": [
+    {
+     "data": {
+      "image/png": "iVBORw0KGgoAAAANSUhEUgAABF0AAAKdCAYAAADx3xbcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
+      "text/plain": [
+       "<matplotlib.figure.Figure at 0x114f46a50>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    }
+   ],
+   "source": [
+    "getActivations(h_conv3,imageToUse)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "The model we used here is of a relatively simple network, but the visualization technique can be extended to give insights into any convolutional network."
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 2",
+   "language": "python",
+   "name": "python2"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.11"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 0
+}