copy edit

Author: cezannec

convolutional-neural-networks/conv-visualization/maxpooling_visualization.ipynb

@@ -23,7 +23,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "import cv2\n",
@@ -57,7 +59,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "import numpy as np\n",
@@ -71,7 +75,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "# Defining four different filters, \n",
@@ -99,15 +105,17 @@
     "\n",
     "In the next cell, we initialize a convolutional layer so that it contains all the created filters. Then add a maxpooling layer, [documented here](http://pytorch.org/docs/stable/_modules/torch/nn/modules/pooling.html), with a kernel size of (2x2) so you can see that the image resolution has been reduced after this step!\n",
     "\n",
-    "A maxpooling layer reduces the x-y size of an input and only keeps the most *active* pixel values. Below is an example of a 2x2 pooling kernel, with a stride of 2, appied to a small patch of grayscale pixel values; reducing the x-y size of the patch by a factor of 2. Only the maximum pixel values in 2x2 remain in the new, pooled output.\n",
+    "A maxpooling layer reduces the x-y size of an input and only keeps the most *active* pixel values. Below is an example of a 2x2 pooling kernel, with a stride of 2, applied to a small patch of grayscale pixel values; reducing the x-y size of the patch by a factor of 2. Only the maximum pixel values in 2x2 remain in the new, pooled output.\n",
     "\n",
     "<img src='notebook_ims/maxpooling_ex.png' height=50% width=50% />"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "import torch\n",
@@ -193,7 +201,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "# plot original image\n",
@@ -233,7 +243,9 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "metadata": {},
+   "metadata": {
+    "collapsed": true
+   },
    "outputs": [],
    "source": [
     "# visualize the output of the pooling layer\n",