Traffic_Sign_Classifier.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Self-Driving Car Engineer Nanodegree\n",
    "\n",
    "## Deep Learning\n",
    "\n",
    "## Project: Build a Traffic Sign Recognition Classifier\n",
    "\n",
    "In this notebook, a template is provided for you to implement your functionality in stages, which is required to successfully complete this project. If additional code is required that cannot be included in the notebook, be sure that the Python code is successfully imported and included in your submission if necessary. \n",
    "\n",
    "> **Note**: Once you have completed all of the code implementations, you need to finalize your work by exporting the iPython Notebook as an HTML document. Before exporting the notebook to html, all of the code cells need to have been run so that reviewers can see the final implementation and output. You can then export the notebook by using the menu above and navigating to  \\n\",\n",
    "    \"**File -> Download as -> HTML (.html)**. Include the finished document along with this notebook as your submission. \n",
    "\n",
    "In addition to implementing code, there is a writeup to complete. The writeup should be completed in a separate file, which can be either a markdown file or a pdf document. There is a [write up template](https://github.com/udacity/CarND-Traffic-Sign-Classifier-Project/blob/master/writeup_template.md) that can be used to guide the writing process. Completing the code template and writeup template will cover all of the [rubric points](https://review.udacity.com/#!/rubrics/481/view) for this project.\n",
    "\n",
    "The [rubric](https://review.udacity.com/#!/rubrics/481/view) contains \"Stand Out Suggestions\" for enhancing the project beyond the minimum requirements. The stand out suggestions are optional. If you decide to pursue the \"stand out suggestions\", you can include the code in this Ipython notebook and also discuss the results in the writeup file.\n",
    "\n",
    "\n",
    ">**Note:** Code and Markdown cells can be executed using the **Shift + Enter** keyboard shortcut. In addition, Markdown cells can be edited by typically double-clicking the cell to enter edit mode."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## Step 0: Load The Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done\n"
     ]
    }
   ],
   "source": [
    "# Load pickled data\n",
    "import pickle\n",
    "import csv\n",
    "\n",
    "training_file = \"train.p\"\n",
    "validation_file= \"valid.p\"\n",
    "testing_file = \"test.p\"\n",
    "\n",
    "with open(training_file, mode='rb') as f:\n",
    "    train = pickle.load(f)\n",
    "with open(validation_file, mode='rb') as f:\n",
    "    valid = pickle.load(f)\n",
    "with open(testing_file, mode='rb') as f:\n",
    "    test = pickle.load(f)\n",
    "    \n",
    "X_train, y_train = train['features'], train['labels']\n",
    "X_valid, y_valid = valid['features'], valid['labels']\n",
    "X_test, y_test = test['features'], test['labels']\n",
    "\n",
    "with open('signnames.csv', 'r') as f:\n",
    "    reader = csv.reader(f)\n",
    "    signClassList = list(reader)\n",
    "print(\"Done\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Step 1: Dataset Summary & Exploration\n",
    "\n",
    "The pickled data is a dictionary with 4 key/value pairs:\n",
    "\n",
    "- `'features'` is a 4D array containing raw pixel data of the traffic sign images, (num examples, width, height, channels).\n",
    "- `'labels'` is a 1D array containing the label/class id of the traffic sign. The file `signnames.csv` contains id -> name mappings for each id.\n",
    "- `'sizes'` is a list containing tuples, (width, height) representing the original width and height the image.\n",
    "- `'coords'` is a list containing tuples, (x1, y1, x2, y2) representing coordinates of a bounding box around the sign in the image. **THESE COORDINATES ASSUME THE ORIGINAL IMAGE. THE PICKLED DATA CONTAINS RESIZED VERSIONS (32 by 32) OF THESE IMAGES**\n",
    "\n",
    "Complete the basic data summary below. Use python, numpy and/or pandas methods to calculate the data summary rather than hard coding the results. For example, the [pandas shape method](http://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.shape.html) might be useful for calculating some of the summary results. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Provide a Basic Summary of the Data Set Using Python, Numpy and/or Pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of training examples = 34799\n",
      "Number of testing examples = 12630\n",
      "Number of validation examples = 4410\n",
      "Image data shape = (32, 32, 3)\n",
      "Number of classes = 43\n"
     ]
    }
   ],
   "source": [
    "### Replace each question mark with the appropriate value. \n",
    "### Use python, pandas or numpy methods rather than hard coding the results\n",
    "import numpy as np\n",
    "\n",
    "# TODO: Number of training examples\n",
    "n_train = len(X_train)\n",
    "\n",
    "# TODO: Number of validation examples\n",
    "n_validation = len(X_valid)\n",
    "\n",
    "# TODO: Number of testing examples.\n",
    "n_test = len(X_test)\n",
    "\n",
    "# TODO: What's the shape of an traffic sign image?\n",
    "image_shape = X_train[0].shape\n",
    "\n",
    "# TODO: How many unique classes/labels there are in the dataset.\n",
    "n_classes = len(np.unique(y_train))\n",
    "\n",
    "print(\"Number of training examples =\", n_train)\n",
    "print(\"Number of testing examples =\", n_test)\n",
    "print(\"Number of validation examples =\", n_validation)\n",
    "print(\"Image data shape =\", image_shape)\n",
    "print(\"Number of classes =\", n_classes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Include an exploratory visualization of the dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualize the German Traffic Signs Dataset using the pickled file(s). This is open ended, suggestions include: plotting traffic sign images, plotting the count of each sign, etc. \n",
    "\n",
    "The [Matplotlib](http://matplotlib.org/) [examples](http://matplotlib.org/examples/index.html) and [gallery](http://matplotlib.org/gallery.html) pages are a great resource for doing visualizations in Python.\n",
    "\n",
    "**NOTE:** It's recommended you start with something simple first. If you wish to do more, come back to it after you've completed the rest of the sections. It can be interesting to look at the distribution of classes in the training, validation and test set. Is the distribution the same? Are there more examples of some classes than others?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY8AAAEWCAYAAACe8xtsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGyFJREFUeJzt3X20XXV95/H3pwHUVhGQwKQJNOgEKzI1YkRmrC0tFgM6\ngl3aglUQ6QRbWMXWtkJnKlbLlJmpD4uqaNQIdJSHERGqWI0oog4IQSIPppSAqUTSJD4BimUMfOeP\nvS8eknNvzib33HNv7vu11ll37+/57X2+Zwfu9/5+++GXqkKSpC5+btQJSJJmHouHJKkzi4ckqTOL\nhySpM4uHJKkzi4ckqTOLh9RHkjlJfpRk/8lsO5lG9bkSWDy0k2h/iY69Hknyk5713+u6v6p6uKqe\nXFXfnsy2XSXZM8n5Sf41yf1J7kjyp8P+XGl7dhl1AtJkqKonjy0nWQf8flV9frz2SXapqi1TkdsO\nOheYA/wycD/wTOBZI81Iwp6HZokkf53kkiQXJXkAeE2S/5jk+iQ/TLIhyblJdm3b75Kkkixs1/93\n+/5nkjyQ5LokB3Rt275/VJJ/TnJfkr9L8tUkrxsn9ecDH6uqH1bVI1W1pqo+Mc7nzk3y6baHckOS\n/57kmq3anpJkbZIfJDm3J6cDk1zb5vTdJB+btIOvnZLFQ7PJK4CPAU8FLgG2AKcDewMvBJYCp0yw\n/auBvwT2Ar4NvL1r2yT7AJcCf9Z+7reAQyfYz/XA3yR5XZJFE389zgN+COwLvB44sU+bo4HnAc+l\nKaAvbuNnA58G9gQWAO/dzmdplrN4aDb5SlX9Q/sX/E+q6saq+lpVbamqu4HlwK9PsP3Hq2pVVf0U\n+Ciw+HG0fRmwuqquaN97F/DdCfbzhzSF7o+ANUnuTHLk1o3aHtOxwFva73Yb8Pd99vc3VXVfVa0D\nrunJ66fAQmBeVf1bVX11gpwki4dmlXt6V5L8cjvM869J7gfeRtMbGM+/9iw/CDx5vIYTtP3F3jyq\neTLp+vF2UlUPVtVfV9UhwNOATwCXJXnqVk33pTk30vsd72Fb4+X1JmBXYFWSW5P067VIj7J4aDbZ\n+hHSHwBuA/59Ve0OvAXIkHPYQDMsBECSAPMH2bCq7gP+huYX/sKt3t4IPNK7b2C/QZOqqg1V9ftV\nNQ84FVjee55G2prFQ7PZU4D7gB8neRYTn++YLJ8CDknyn5PsQnPOZe54jZOclWRJkt2SPJFm+Or7\nwJ297dohsE8Cf5XkSUmeDbxm0KSS/E6SsSL2Q5pC+3CXL6bZxeKh2exNNCeVH6DphVwy7A+sqo3A\n7wLvBL4HPAO4GXhogs0uaNveCxwOvLSqHuzT7g9ohrY2Ah8BLtrOfnu9ALgxyY9phsZO9f4RTSRO\nBiWNTpI5NEXhlVX15Une9zuAParq5MncrwT2PKQpl2RpkqcmeQLN5bxbgBsmYb8HJfkPaRwGnARc\nvqP7lfrxDnNp6v0qzeW7uwG3A8dW1aDDSxPZvd3vPJqhq3Oq6lOTsF9pGw5bSZI6c9hKktTZTjts\ntffee9fChQtHnYYkzRg33XTTd6tq3EvHe+20xWPhwoWsWrVq1GlI0oyR5F8GbeuwlSSpM4uHJKkz\ni4ckqTOLhySpM4uHJKkzi4ckqTOLhySps6EVjyT7JflikjVJbk9yehvfK8nKdjrNlUn2bONJcm6S\ntUluSXJIz75ObNvf6QxnkjR6w+x5bAHeVFXPAg4DTk1yEHAGcHVVLQKubtcBjgIWta9lwHnQFBvg\nLJr5Bg4FzhorOJKk0RjaHeZVtYFmyk2q6oEka2im2zyGZkIbaCa5uQZ4cxu/sJ3T+fokeySZ17Zd\nWVXfB0iyElhKM9HNtLLwjE/3ja8756VTnIkkDdeUnPNIshB4LvA1YN+2sIwVmH3aZvOBe3o2W9/G\nxov3+5xlSVYlWbV58+bJ/AqSpB5DLx5JngxcBryxqu6fqGmfWE0Q3zZYtbyqllTVkrlzB3q2lyTp\ncRjqgxGT7EpTOD5aVZ9owxuTzKuqDe2w1KY2vh7Yr2fzBTTTc67nZ8NcY/Frhpn3RByakqThXm0V\n4MPAmqp6Z89bVwJjV0ydCFzREz+hZwrN+9phrc8CRybZsz1RfmQbkySNyDB7Hi8EXgvcmmR1G/sL\n4Bzg0iQnA98GXtW+dxVwNLAWeJBm/mWq6vtJ3g7c2LZ729jJc0nSaAzzaquv0P98BcARfdoXcOo4\n+1oBrJi87CRJO8I7zCVJnVk8JEmd7bTT0Grm8Ao2aeax5yFJ6sziIUnqzGErTQmHpqSdiz0PSVJn\nFg9JUmcOW0nSJJhtQ7P2PCRJnVk8JEmdWTwkSZ1ZPCRJnVk8JEmdebWVOpltV5RI6s+ehySps2FO\nQ7siyaYkt/XELkmyun2tG5thMMnCJD/pee/9Pds8L8mtSdYmObed3laSNELDHLY6H3gPcOFYoKp+\nd2w5yTuA+3ra31VVi/vs5zxgGXA9zVS1S4HPDCHfkXEoSNJMM7SeR1VdC/Sda7ztPfwOcNFE+0gy\nD9i9qq5rp6m9EDh2snOVJHUzqnMeLwI2VtWdPbEDktyc5EtJXtTG5gPre9qsb2OSpBEa1dVWx/PY\nXscGYP+q+l6S5wGfTPJsoN/5jRpvp0mW0Qxxsf/++09iupKkXlPe80iyC/DbwCVjsap6qKq+1y7f\nBNwFHEjT01jQs/kC4N7x9l1Vy6tqSVUtmTt37jDSlyQxmmGrFwP/VFWPDkclmZtkTrv8dGARcHdV\nbQAeSHJYe57kBOCKEeQsSeoxzEt1LwKuA56ZZH2Sk9u3jmPbE+W/BtyS5BvAx4E3VNXYyfY/AD4E\nrKXpkexUV1pJ0kw0tHMeVXX8OPHX9YldBlw2TvtVwMGTmpwkaYd4h7kkqTOLhySpM4uHJKkzi4ck\nqTOLhySpM+fzmOZ8aKKk6ciehySpM4uHJKkzi4ckqTOLhySpM4uHJKkzi4ckqTOLhySpM4uHJKkz\ni4ckqTOLhySpM4uHJKmzYU5DuyLJpiS39cTemuQ7SVa3r6N73jszydokdyR5SU98aRtbm+SMYeUr\nSRrcMHse5wNL+8TfVVWL29dVAEkOopnb/NntNu9LMifJHOC9wFHAQcDxbVtJ0ggNcw7za5MsHLD5\nMcDFVfUQ8K0ka4FD2/fWVtXdAEkubtt+c5LTlSR1MIpzHqcluaUd1tqzjc0H7ulps76NjRfvK8my\nJKuSrNq8efNk5y1Jak118TgPeAawGNgAvKONp0/bmiDeV1Utr6olVbVk7ty5O5qrJGkcUzoZVFVt\nHFtO8kHgU+3qemC/nqYLgHvb5fHikqQRmdKeR5J5PauvAMauxLoSOC7JE5IcACwCbgBuBBYlOSDJ\nbjQn1a+cypwlSdsaWs8jyUXA4cDeSdYDZwGHJ1lMM/S0DjgFoKpuT3IpzYnwLcCpVfVwu5/TgM8C\nc4AVVXX7sHKWJA1mmFdbHd8n/OEJ2p8NnN0nfhVw1SSmJknaQd5hLknqzOIhSerM4iFJ6sziIUnq\nzOIhSerM4iFJ6sziIUnqzOIhSerM4iFJ6sziIUnqbEqfqitp6iw849PbxNad89IRZKKdkT0PSVJn\nFg9JUmcOW2kbDndI2h57HpKkziwekqTOHLbaiTn8JGlYhtbzSLIiyaYkt/XE/leSf0pyS5LLk+zR\nxhcm+UmS1e3r/T3bPC/JrUnWJjk3SYaVsyRpMMMctjofWLpVbCVwcFX9CvDPwJk9791VVYvb1xt6\n4ucBy4BF7WvrfUqSpthAw1ZJTgc+AjwAfAh4LnBGVX1uvG2q6tokC7eK9ba/Hnjldj53HrB7VV3X\nrl8IHAt8ZpC8p5N+Q0gwumEkh7Qk7YhBex6vr6r7gSOBucBJwDk7+Nmv57FF4IAkNyf5UpIXtbH5\nwPqeNuvbWF9JliVZlWTV5s2bdzA9SdJ4Bi0eY+cZjgY+UlXf6Il1luS/AluAj7ahDcD+VfVc4E+A\njyXZfZzPqPH2W1XLq2pJVS2ZO3fu401PkrQdg15tdVOSzwEHAGcmeQrwyOP5wCQnAi8DjqiqAqiq\nh4CH2uWbktwFHEjT01jQs/kC4N7H87mSpMkzaM/jZOAM4PlV9SCwG83QVSdJlgJvBl7e7mcsPjfJ\nnHb56TQnxu+uqg3AA0kOa6+yOgG4ouvnSpIm16DFo4CDgD9q138BeOJEGyS5CLgOeGaS9UlOBt4D\nPAVYudUlub8G3JLkG8DHgTdU1ffb9/6A5iT9WuAuZuDJckna2Qw6bPU+mmGq3wTeRnPV1WXA88fb\noKqO7xP+8DhtL2v31++9VcDBA+YpSZoCgxaPF1TVIUluBqiqHyTZbYh5SZKmsUGHrX7anpMoaM5R\n8DhPmEuSZr5Bex7nApcD+yQ5m+bmvv82tKw0I023GyElDc9AxaOqPprkJuAImnsvjq2qNUPNTJI0\nbU1YPJLs1bO6Cbio972eK6IkSbPI9noeN9Gc5xjvTu+nT3pGkqRpb8LiUVUHTFUikqSZY+DJoJL8\nNvCrND2OL1fVJ4eWlSRpWhvoUt0k7wPeANwK3Aa8Icl7h5mYJGn6GrTn8es0kziN3edxAU0hkSTN\nQoPeJHgHsH/P+n7ALZOfjiRpJhi05/E0YE2SG9r15wPXJbkSoKpePozkJEnT06DF4y1DzUKaRpyi\nd+fnv/GOG/QO8y8BtLP77dIT9yZBSZqFBioeSZYBbwd+QvNAxOBNgpI0aw06bPVnwLOr6rvDTEaS\nNDMMerXVXcCD220lSZoVBi0eZwL/N8kHkpw79treRklWJNmU5Lae2F5JVia5s/25ZxtPu9+1SW5J\nckjPNie27e9McmLXLylJmlyDFo8PAF8Arqd5WOLYa3vOB5ZuFTsDuLqqFgFXt+sARwGL2tcy4Dx4\n9Mm+ZwEvAA4FzhorOJKk0Rj0nMeWqvqTrjuvqmuTLNwqfAxweLt8AXAN8OY2fmF7F/v1SfZIMq9t\nu3Lsyq4kK2kK0kVIkkZi0J7HF5MsSzKvHXbaa6u5PrrYt6o2ALQ/92nj84F7etqtb2PjxbfR5rgq\nyarNmzc/zvQkSdszaM/j1e3PM3tik32p7nhzhowX3zZYtRxYDrBkyZK+bSTwJjFpRw16k+Bkzuux\nMcm8qtrQDkttauPraZ6ZNWYBcG8bP3yr+DWTmI8kqaNBh61IcnCS30lywtjrcX7mlcDYFVMnAlf0\nxE9or7o6DLivHdb6LHBkkj3bE+VHtjFJ0ogMeof5WTR//R8EXEVzZdRXgAu3s91F7XZ7J1lPc9XU\nOcClSU4Gvg28qm1+FXA0sJbmnpKToHkESpK3Aze27d7mY1EkabQGPefxSuA5wM1VdVKSfYEPbW+j\nqjp+nLeO6NO2gFPH2c8KYMWAuUqShmzQYaufVNUjwJb24Yib8LlWkjRrDdrzWJVkD+CDNDcH/gi4\nYeJNJEk7q0GvtvrDdvH9Sf4R2L2qnElQkmapgYat2pPbAFTVOuD29iS6JGkWGvScxxFJrmrvMD+Y\n5hlXTxliXpKkaWzQYatXJ/ld4Faay2iPr6qvDjUzSdK0Neiw1SLgdOAyYB3w2iQ/P8S8JEnT2KDD\nVv8A/GVVnQL8OnAnP7tpT5I0ywx6qe6hVXU/PHoz3zuSXDm8tCRJ09mEPY8kfw5QVfcnedVWb580\ntKwkSdPa9oatjutZPnOr97aeIVCSNEtsr3hknOV+65KkWWJ7xaPGWe63LkmaJbZ3wvw5Se6n6WU8\nqV2mXX/iUDOTJE1bExaPqpozVYlIkmaOgWcSlCRpzJQXjyTPTLK653V/kjcmeWuS7/TEj+7Z5swk\na5PckeQlU52zJOmxBr1JcNJU1R3AYoAkc4DvAJfT3Dfyrqr62972SQ6iuWT42cAvAp9PcmBVPTyl\niUuSHjXqYasjgLuq6l8maHMMcHFVPVRV36KZ4/zQKclOktTXqIvHccBFPeunJbklyYoke7ax+cA9\nPW3WtzFJ0oiMrHgk2Q14OfB/2tB5wDNohrQ2AO8Ya9pn8773mCRZlmRVklWbN2+e5IwlSWNG2fM4\nCvh6VW0EqKqNVfVwVT1CM1f62NDUemC/nu0WAPf222FVLa+qJVW1ZO7cuUNMXZJmt1EWj+PpGbJK\nMq/nvVcAt7XLVwLHJXlCkgOARcANU5alJGkbU361FUA7kdRvAaf0hP9nksU0Q1Lrxt6rqtuTXAp8\nE9gCnOqVVpI0WiMpHlX1IPC0rWKvnaD92cDZw85LkjSYUV9tJUmagUbS85AGtfCMT/eNrzvnpVOc\niYbFf+OZyZ6HJKkzi4ckqTOLhySpM4uHJKkzi4ckqTOLhySpM4uHJKkzi4ckqTNvEpQ0I3lz4WjZ\n85AkdWbxkCR1ZvGQJHVm8ZAkdWbxkCR15tVWmrG82ubx63fsPG7qYmQ9jyTrktyaZHWSVW1sryQr\nk9zZ/tyzjSfJuUnWJrklySGjyluSNPphq9+oqsVVtaRdPwO4uqoWAVe36wBHAYva1zLgvCnPVJL0\nqFEXj60dA1zQLl8AHNsTv7Aa1wN7JJk3igQlSaMtHgV8LslNSZa1sX2ragNA+3OfNj4fuKdn2/Vt\n7DGSLEuyKsmqzZs3DzF1SZrdRnnC/IVVdW+SfYCVSf5pgrbpE6ttAlXLgeUAS5Ys2eZ9SdLkGFnx\nqKp725+bklwOHApsTDKvqja0w1Kb2ubrgf16Nl8A3DulCUvyCjc9aiTDVkl+IclTxpaBI4HbgCuB\nE9tmJwJXtMtXAie0V10dBtw3NrwlSZp6o+p57AtcnmQsh49V1T8muRG4NMnJwLeBV7XtrwKOBtYC\nDwInTX3KkqQxIykeVXU38Jw+8e8BR/SJF3DqFKQmTSsOE2m6mm6X6kqSZgCLhySpM4uHJKkzi4ck\nqTOLhySpM4uHJKkz5/OQNHRecrzzsechSerM4iFJ6sxhqxnMoYCJOdXq7Oa//3DZ85AkdWbxkCR1\n5rCVNGIOP04vDncNxp6HJKkzi4ckqTOHraRJ4vCTHo+Z+t+NPQ9JUmdTXjyS7Jfki0nWJLk9yelt\n/K1JvpNkdfs6umebM5OsTXJHkpdMdc6SpMcaxbDVFuBNVfX1JE8Bbkqysn3vXVX1t72NkxwEHAc8\nG/hF4PNJDqyqh6c0a2kHzNShCU2OnfHff8p7HlW1oaq+3i4/AKwB5k+wyTHAxVX1UFV9C1gLHDr8\nTCVJ4xnpOY8kC4HnAl9rQ6cluSXJiiR7trH5wD09m61nnGKTZFmSVUlWbd68eUhZS5JGdrVVkicD\nlwFvrKr7k5wHvB2o9uc7gNcD6bN59dtnVS0HlgMsWbKkbxtpZzHdhkKmWz4arpH0PJLsSlM4PlpV\nnwCoqo1V9XBVPQJ8kJ8NTa0H9uvZfAFw71TmK0l6rFFcbRXgw8CaqnpnT3xeT7NXALe1y1cCxyV5\nQpIDgEXADVOVryRpW6MYtnoh8Frg1iSr29hfAMcnWUwzJLUOOAWgqm5PcinwTZortU71SitpeKbT\ns51m+1DYRN9/1MdmyotHVX2F/ucxrppgm7OBs4eWlCSpE+8wlyR15rOt+phO3XZNL6MeKtBo+e//\nM/Y8JEmdWTwkSZ1ZPCRJnVk8JEmdWTwkSZ1ZPCRJnVk8JEmdWTwkSZ1ZPCRJnVk8JEmdWTwkSZ1Z\nPCRJnVk8JEmdWTwkSZ1ZPCRJnc2Y4pFkaZI7kqxNcsao85Gk2WxGFI8kc4D3AkcBB9HMd37QaLOS\npNlrRhQP4FBgbVXdXVX/D7gYOGbEOUnSrJWqGnUO25XklcDSqvr9dv21wAuq6rSt2i0DlrWrzwTu\nmISP3xv47iTsZ2fksRmfx2Z8HpvxjfrY/FJVzR2k4UyZwzx9YttUvapaDiyf1A9OVlXVksnc587C\nYzM+j834PDbjm0nHZqYMW60H9utZXwDcO6JcJGnWmynF40ZgUZIDkuwGHAdcOeKcJGnWmhHDVlW1\nJclpwGeBOcCKqrp9ij5+UofBdjIem/F5bMbnsRnfjDk2M+KEuSRpepkpw1aSpGnE4iFJ6sziMQ4f\nh/JYSVYk2ZTktp7YXklWJrmz/bnnKHMchST7JflikjVJbk9yehv32CRPTHJDkm+0x+av2vgBSb7W\nHptL2otgZqUkc5LcnORT7fqMOTYWjz58HEpf5wNLt4qdAVxdVYuAq9v12WYL8KaqehZwGHBq+9+K\nxwYeAn6zqp4DLAaWJjkM+B/Au9pj8wPg5BHmOGqnA2t61mfMsbF49OfjULZSVdcC398qfAxwQbt8\nAXDslCY1DVTVhqr6erv8AM0vgvl4bKjGj9rVXdtXAb8JfLyNz8pjA5BkAfBS4EPtephBx8bi0d98\n4J6e9fVtTI+1b1VtgOaXKLDPiPMZqSQLgecCX8NjAzw6LLMa2ASsBO4CflhVW9oms/n/rXcDfw48\n0q4/jRl0bCwe/Q30OBRpTJInA5cBb6yq+0edz3RRVQ9X1WKap0IcCjyrX7OpzWr0krwM2FRVN/WG\n+zSdtsdmRtwkOAI+DmUwG5PMq6oNSebR/HU56yTZlaZwfLSqPtGGPTY9quqHSa6hOS+0R5Jd2r+w\nZ+v/Wy8EXp7kaOCJwO40PZEZc2zsefTn41AGcyVwYrt8InDFCHMZiXac+sPAmqp6Z89bHptkbpI9\n2uUnAS+mOSf0ReCVbbNZeWyq6syqWlBVC2l+v3yhqn6PGXRsvMN8HO1fBO/mZ49DOXvEKY1UkouA\nw2keGb0ROAv4JHApsD/wbeBVVbX1SfWdWpJfBb4M3MrPxq7/gua8x2w/Nr9Cc9J3Ds0fqpdW1duS\nPJ3mIpS9gJuB11TVQ6PLdLSSHA78aVW9bCYdG4uHJKkzh60kSZ1ZPCRJnVk8JEmdWTwkSZ1ZPCRJ\nnVk8pB2U5N8luTjJXUm+meSqJAf2PoFY2tl4h7m0A9qbBC8HLqiq49rYYmDfkSYmDZk9D2nH/Abw\n06p6/1igqlbT82DNJAuTfDnJ19vXf2rj85Jcm2R1ktuSvKh9kOD57fqtSf546r+StH32PKQdczBw\n03babAJ+q6r+Lcki4CJgCfBq4LNVdXY7h8zP08x7Mb+qDgYYe7yHNN1YPKTh2xV4Tzuc9TBwYBu/\nEVjRPljxk1W1OsndwNOT/B3waeBzI8lY2g6HraQdczvwvO20+WOa54E9h6bHsRs8OsHWrwHfAf4+\nyQlV9YO23TXAqbQTBUnTjcVD2jFfAJ6Q5L+MBZI8H/ilnjZPBTZU1SPAa2keFEiSX6KZ0+GDNE/m\nPSTJ3sDPVdVlwF8Ch0zN15C6cdhK2gFVVUleAbw7yRnAvwHrgDf2NHsfcFmSV9E8cvvHbfxw4M+S\n/BT4EXACzcxxH0ky9ofdmUP/EtLj4FN1JUmdOWwlSerM4iFJ6sziIUnqzOIhSerM4iFJ6sziIUnq\nzOIhSers/wOUW2iKidb9CgAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f658139048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f658f20eb8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f658f18860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Data exploration visualization code goes here.\n",
    "### Feel free to use as many code cells as needed.\n",
    "import matplotlib.pyplot as plt\n",
    "# Visualizations will be shown in the notebook.\n",
    "y_switch = [y_train,y_valid,y_test,[\"Training\",\"Validation\",\"Testing\"]]\n",
    "\n",
    "for count in range(0,3):\n",
    "    x, y= np.unique(y_switch[count], return_counts=True)\n",
    "    plt.bar(x,y)\n",
    "    plt.title(y_switch[3][count] +' Signs')\n",
    "    plt.xlabel('Class')\n",
    "    plt.ylabel('Examples')\n",
    "    plt.show()\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "----\n",
    "\n",
    "## Step 2: Design and Test a Model Architecture\n",
    "\n",
    "Design and implement a deep learning model that learns to recognize traffic signs. Train and test your model on the [German Traffic Sign Dataset](http://benchmark.ini.rub.de/?section=gtsrb&subsection=dataset).\n",
    "\n",
    "The LeNet-5 implementation shown in the [classroom](https://classroom.udacity.com/nanodegrees/nd013/parts/fbf77062-5703-404e-b60c-95b78b2f3f9e/modules/6df7ae49-c61c-4bb2-a23e-6527e69209ec/lessons/601ae704-1035-4287-8b11-e2c2716217ad/concepts/d4aca031-508f-4e0b-b493-e7b706120f81) at the end of the CNN lesson is a solid starting point. You'll have to change the number of classes and possibly the preprocessing, but aside from that it's plug and play! \n",
    "\n",
    "With the LeNet-5 solution from the lecture, you should expect a validation set accuracy of about 0.89. To meet specifications, the validation set accuracy will need to be at least 0.93. It is possible to get an even higher accuracy, but 0.93 is the minimum for a successful project submission. \n",
    "\n",
    "There are various aspects to consider when thinking about this problem:\n",
    "\n",
    "- Neural network architecture (is the network over or underfitting?)\n",
    "- Play around preprocessing techniques (normalization, rgb to grayscale, etc)\n",
    "- Number of examples per label (some have more than others).\n",
    "- Generate fake data.\n",
    "\n",
    "Here is an example of a [published baseline model on this problem](http://yann.lecun.com/exdb/publis/pdf/sermanet-ijcnn-11.pdf). It's not required to be familiar with the approach used in the paper but, it's good practice to try to read papers like these."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pre-process the Data Set (normalization, grayscale, etc.)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Minimally, the image data should be normalized so that the data has mean zero and equal variance. For image data, `(pixel - 128)/ 128` is a quick way to approximately normalize the data and can be used in this project. \n",
    "\n",
    "Other pre-processing steps are optional. You can try different techniques to see if it improves performance. \n",
    "\n",
    "Use the code cell (or multiple code cells, if necessary) to implement the first step of your project."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "### Preprocess the data here. It is required to normalize the data. Other preprocessing steps could include \n",
    "### converting to grayscale, etc.\n",
    "### Feel free to use as many code cells as needed.\n",
    "\n",
    "xTrainGray = np.sum(X_train/3, axis=3, keepdims=True)\n",
    "xTrainNormalize = (xTrainGray - 128.)/128.\n",
    "\n",
    "xValidGray = np.sum(X_valid/3, axis=3, keepdims=True)\n",
    "xValidNormalize = (xValidGray - 128.)/128.\n",
    "\n",
    "xTestGray = np.sum(X_test/3, axis=3, keepdims=True)\n",
    "xTestNormalize = (xTestGray - 128.)/128.\n",
    "\n",
    "X_train = xTrainNormalize\n",
    "X_valid = xValidNormalize\n",
    "X_test = xTestNormalize"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Architecture - Modified LeNet 5 - http://yann.lecun.com/exdb/lenet/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow.contrib.layers import flatten\n",
    "from sklearn.utils import shuffle"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Shuffled\n"
     ]
    }
   ],
   "source": [
    "X_train, y_train = shuffle(X_train, y_train)\n",
    "print(\"Shuffled\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Paramaters\n",
    "'''\n",
    "#96.4%\n",
    "EPOCHS = 50\n",
    "BATCH_SIZE = 1000\n",
    "RATE = 0.002\n",
    "KEEP_PROB = 0.5\n",
    "'''\n",
    "\n",
    "EPOCHS = 100\n",
    "BATCH_SIZE = 1000\n",
    "LEARN_RATE = 0.002\n",
    "KEEP_PROB = 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def LeDaveNet(conv):\n",
    "    \n",
    "    MEAN = 0\n",
    "    STDDEV = 0.1\n",
    "    \n",
    "    #Convolution 1: In = (32,32,1) Out = (28,28,6)\n",
    "    weight1 = tf.Variable(tf.truncated_normal(shape=(5, 5, 1, 6), mean = MEAN, stddev = STDDEV))\n",
    "    bias1 = tf.Variable(tf.zeros(6))\n",
    "    conv = tf.nn.conv2d(conv, weight1, strides=[1, 1, 1, 1], padding='VALID')\n",
    "    conv = tf.nn.bias_add(conv, bias1)\n",
    "    \n",
    "    #Activation:\n",
    "    conv = tf.nn.relu(conv)\n",
    "    \n",
    "    #Max Pooling: In = (28,28,6) Out = (14,14,6)\n",
    "    conv = tf.nn.max_pool(conv, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')\n",
    "    \n",
    "    #Convolution 2: In = (14,14,6) Out = (10,10,16)\n",
    "    weight2 = tf.Variable(tf.truncated_normal(shape=(5, 5, 6, 16), mean = MEAN, stddev = STDDEV))\n",
    "    bias2 = tf.Variable(tf.zeros(16))\n",
    "    conv = tf.nn.conv2d(conv, weight2, strides=[1, 1, 1, 1], padding='VALID')\n",
    "    conv = tf.nn.bias_add(conv, bias2)\n",
    "    \n",
    "    #Activation:\n",
    "    conv = tf.nn.relu(conv)\n",
    "    \n",
    "    #Max Pooling: In = (10,10,16) Out = (5,5,16)\n",
    "    conv = tf.nn.max_pool(conv, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')\n",
    "    \n",
    "    #Flatten: In = (5,5,16) Out = (400)\n",
    "    conv = flatten(conv)\n",
    "    \n",
    "    #Convolution 3, flat: In = (400) Out = (120)\n",
    "    weight3 = tf.Variable(tf.truncated_normal(shape=(400, 120), mean = MEAN, stddev = STDDEV))\n",
    "    bias3 = tf.Variable(tf.zeros(120))    \n",
    "    conv = tf.add(tf.matmul(conv, weight3), bias3)\n",
    "    \n",
    "    #Activation:\n",
    "    conv = tf.nn.relu(conv)\n",
    "    \n",
    "    #Dropout:\n",
    "    conv = tf.nn.dropout(conv, keepProb)\n",
    "    \n",
    "    #Convolution 4, flat: In = (120) Out = (84)\n",
    "    weight4 = tf.Variable(tf.truncated_normal(shape=(120, 84), mean = MEAN, stddev = STDDEV))\n",
    "    bias4 = tf.Variable(tf.zeros(84)) \n",
    "    conv = tf.add(tf.matmul(conv, weight4), bias4)\n",
    "    \n",
    "    #Activation:\n",
    "    conv = tf.nn.relu(conv)\n",
    "    \n",
    "    #Dropout:\n",
    "    conv = tf.nn.dropout(conv, keepProb)\n",
    "    \n",
    "    #Convolution 5, flat: In = (84) Out = (43)\n",
    "    weight5 = tf.Variable(tf.truncated_normal(shape=(84, 43), mean = MEAN, stddev = STDDEV))\n",
    "    bias5 = tf.Variable(tf.zeros(43)) \n",
    "    logits = tf.add(tf.matmul(conv, weight5), bias5)\n",
    "    \n",
    "    return logits"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "x = tf.placeholder(tf.float32, (None, 32, 32, 1))\n",
    "y = tf.placeholder(tf.int32, (None))\n",
    "keepProb = tf.placeholder(tf.float32)\n",
    "oneHotY = tf.one_hot(y, 43)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "logits = LeDaveNet(x)\n",
    "crossEntropy = tf.nn.softmax_cross_entropy_with_logits(logits = logits, labels = oneHotY)\n",
    "lossOperation = tf.reduce_mean(crossEntropy)\n",
    "optimizer = tf.train.AdamOptimizer(learning_rate = LEARN_RATE)\n",
    "trainingOperation = optimizer.minimize(lossOperation)\n",
    "correctPrediction = tf.equal(tf.argmax(logits, 1), tf.argmax(oneHotY, 1))\n",
    "accuracyOperation = tf.reduce_mean(tf.cast(correctPrediction, tf.float32))\n",
    "saver = tf.train.Saver()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "##All evaluation code referenced from Lesson 8, Part36 solution: \n",
    "##https://github.com/udacity/CarND-LeNet-Lab/blob/master/LeNet-Lab-Solution.ipynb\n",
    "\n",
    "def evaluate(xData, yData):\n",
    "    numExamples = len(xData)\n",
    "    totalAccuracy = 0\n",
    "    sess = tf.get_default_session()\n",
    "    for offset in range(0, numExamples, BATCH_SIZE):\n",
    "        xBatch, yBatch = xData[offset:offset+BATCH_SIZE], yData[offset:offset+BATCH_SIZE]\n",
    "        accuracy = sess.run(accuracyOperation, feed_dict={x: xBatch, y: yBatch, keepProb: 1.0})\n",
    "        totalAccuracy += (accuracy * len(xBatch))\n",
    "    return totalAccuracy / numExamples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training...\n",
      "\n",
      "EPOCH 1 ...\n",
      "Validation Accuracy = 21.156%\n",
      "\n",
      "EPOCH 2 ...\n",
      "Validation Accuracy = 49.615%\n",
      "\n",
      "EPOCH 3 ...\n",
      "Validation Accuracy = 68.367%\n",
      "\n",
      "EPOCH 4 ...\n",
      "Validation Accuracy = 78.095%\n",
      "\n",
      "EPOCH 5 ...\n",
      "Validation Accuracy = 82.766%\n",
      "\n",
      "EPOCH 6 ...\n",
      "Validation Accuracy = 85.147%\n",
      "\n",
      "EPOCH 7 ...\n",
      "Validation Accuracy = 86.531%\n",
      "\n",
      "EPOCH 8 ...\n",
      "Validation Accuracy = 87.778%\n",
      "\n",
      "EPOCH 9 ...\n",
      "Validation Accuracy = 87.846%\n",
      "\n",
      "EPOCH 10 ...\n",
      "Validation Accuracy = 90.023%\n",
      "\n",
      "EPOCH 11 ...\n",
      "Validation Accuracy = 90.612%\n",
      "\n",
      "EPOCH 12 ...\n",
      "Validation Accuracy = 91.134%\n",
      "\n",
      "EPOCH 13 ...\n",
      "Validation Accuracy = 90.816%\n",
      "\n",
      "EPOCH 14 ...\n",
      "Validation Accuracy = 91.746%\n",
      "\n",
      "EPOCH 15 ...\n",
      "Validation Accuracy = 92.177%\n",
      "\n",
      "EPOCH 16 ...\n",
      "Validation Accuracy = 93.379%\n",
      "\n",
      "EPOCH 17 ...\n",
      "Validation Accuracy = 93.424%\n",
      "\n",
      "EPOCH 18 ...\n",
      "Validation Accuracy = 93.356%\n",
      "\n",
      "EPOCH 19 ...\n",
      "Validation Accuracy = 93.696%\n",
      "\n",
      "EPOCH 20 ...\n",
      "Validation Accuracy = 93.288%\n",
      "\n",
      "EPOCH 21 ...\n",
      "Validation Accuracy = 93.878%\n",
      "\n",
      "EPOCH 22 ...\n",
      "Validation Accuracy = 94.104%\n",
      "\n",
      "EPOCH 23 ...\n",
      "Validation Accuracy = 93.719%\n",
      "\n",
      "EPOCH 24 ...\n",
      "Validation Accuracy = 94.127%\n",
      "\n",
      "EPOCH 25 ...\n",
      "Validation Accuracy = 94.512%\n",
      "\n",
      "EPOCH 26 ...\n",
      "Validation Accuracy = 94.399%\n",
      "\n",
      "EPOCH 27 ...\n",
      "Validation Accuracy = 94.649%\n",
      "\n",
      "EPOCH 28 ...\n",
      "Validation Accuracy = 94.785%\n",
      "\n",
      "EPOCH 29 ...\n",
      "Validation Accuracy = 94.989%\n",
      "\n",
      "EPOCH 30 ...\n",
      "Validation Accuracy = 95.057%\n",
      "\n",
      "EPOCH 31 ...\n",
      "Validation Accuracy = 94.399%\n",
      "\n",
      "EPOCH 32 ...\n",
      "Validation Accuracy = 94.739%\n",
      "\n",
      "EPOCH 33 ...\n",
      "Validation Accuracy = 95.215%\n",
      "\n",
      "EPOCH 34 ...\n",
      "Validation Accuracy = 94.762%\n",
      "\n",
      "EPOCH 35 ...\n",
      "Validation Accuracy = 95.125%\n",
      "\n",
      "EPOCH 36 ...\n",
      "Validation Accuracy = 94.785%\n",
      "\n",
      "EPOCH 37 ...\n",
      "Validation Accuracy = 94.943%\n",
      "\n",
      "EPOCH 38 ...\n",
      "Validation Accuracy = 94.943%\n",
      "\n",
      "EPOCH 39 ...\n",
      "Validation Accuracy = 94.762%\n",
      "\n",
      "EPOCH 40 ...\n",
      "Validation Accuracy = 94.649%\n",
      "\n",
      "EPOCH 41 ...\n",
      "Validation Accuracy = 94.966%\n",
      "\n",
      "EPOCH 42 ...\n",
      "Validation Accuracy = 95.170%\n",
      "\n",
      "EPOCH 43 ...\n",
      "Validation Accuracy = 94.921%\n",
      "\n",
      "EPOCH 44 ...\n",
      "Validation Accuracy = 95.442%\n",
      "\n",
      "EPOCH 45 ...\n",
      "Validation Accuracy = 95.261%\n",
      "\n",
      "EPOCH 46 ...\n",
      "Validation Accuracy = 95.283%\n",
      "\n",
      "EPOCH 47 ...\n",
      "Validation Accuracy = 95.215%\n",
      "\n",
      "EPOCH 48 ...\n",
      "Validation Accuracy = 95.873%\n",
      "\n",
      "EPOCH 49 ...\n",
      "Validation Accuracy = 95.261%\n",
      "\n",
      "EPOCH 50 ...\n",
      "Validation Accuracy = 95.283%\n",
      "\n",
      "EPOCH 51 ...\n",
      "Validation Accuracy = 95.057%\n",
      "\n",
      "EPOCH 52 ...\n",
      "Validation Accuracy = 95.737%\n",
      "\n",
      "EPOCH 53 ...\n",
      "Validation Accuracy = 95.578%\n",
      "\n",
      "EPOCH 54 ...\n",
      "Validation Accuracy = 95.420%\n",
      "\n",
      "EPOCH 55 ...\n",
      "Validation Accuracy = 95.805%\n",
      "\n",
      "EPOCH 56 ...\n",
      "Validation Accuracy = 95.556%\n",
      "\n",
      "EPOCH 57 ...\n",
      "Validation Accuracy = 95.102%\n",
      "\n",
      "EPOCH 58 ...\n",
      "Validation Accuracy = 96.032%\n",
      "\n",
      "EPOCH 59 ...\n",
      "Validation Accuracy = 95.737%\n",
      "\n",
      "EPOCH 60 ...\n",
      "Validation Accuracy = 96.145%\n",
      "\n",
      "EPOCH 61 ...\n",
      "Validation Accuracy = 94.943%\n",
      "\n",
      "EPOCH 62 ...\n",
      "Validation Accuracy = 95.578%\n",
      "\n",
      "EPOCH 63 ...\n",
      "Validation Accuracy = 96.100%\n",
      "\n",
      "EPOCH 64 ...\n",
      "Validation Accuracy = 95.850%\n",
      "\n",
      "EPOCH 65 ...\n",
      "Validation Accuracy = 95.510%\n",
      "\n",
      "EPOCH 66 ...\n",
      "Validation Accuracy = 95.306%\n",
      "\n",
      "EPOCH 67 ...\n",
      "Validation Accuracy = 95.578%\n",
      "\n",
      "EPOCH 68 ...\n",
      "Validation Accuracy = 95.624%\n",
      "\n",
      "EPOCH 69 ...\n",
      "Validation Accuracy = 95.669%\n",
      "\n",
      "EPOCH 70 ...\n",
      "Validation Accuracy = 95.737%\n",
      "\n",
      "EPOCH 71 ...\n",
      "Validation Accuracy = 96.032%\n",
      "\n",
      "EPOCH 72 ...\n",
      "Validation Accuracy = 95.760%\n",
      "\n",
      "EPOCH 73 ...\n",
      "Validation Accuracy = 95.578%\n",
      "\n",
      "EPOCH 74 ...\n",
      "Validation Accuracy = 96.168%\n",
      "\n",
      "EPOCH 75 ...\n",
      "Validation Accuracy = 95.125%\n",
      "\n",
      "EPOCH 76 ...\n",
      "Validation Accuracy = 95.986%\n",
      "\n",
      "EPOCH 77 ...\n",
      "Validation Accuracy = 96.372%\n",
      "\n",
      "EPOCH 78 ...\n",
      "Validation Accuracy = 95.737%\n",
      "\n",
      "EPOCH 79 ...\n",
      "Validation Accuracy = 95.873%\n",
      "\n",
      "EPOCH 80 ...\n",
      "Validation Accuracy = 95.941%\n",
      "\n",
      "EPOCH 81 ...\n",
      "Validation Accuracy = 95.805%\n",
      "\n",
      "EPOCH 82 ...\n",
      "Validation Accuracy = 96.077%\n",
      "\n",
      "EPOCH 83 ...\n",
      "Validation Accuracy = 95.692%\n",
      "\n",
      "EPOCH 84 ...\n",
      "Validation Accuracy = 96.054%\n",
      "\n",
      "EPOCH 85 ...\n",
      "Validation Accuracy = 95.850%\n",
      "\n",
      "EPOCH 86 ...\n",
      "Validation Accuracy = 95.873%\n",
      "\n",
      "EPOCH 87 ...\n",
      "Validation Accuracy = 96.327%\n",
      "\n",
      "EPOCH 88 ...\n",
      "Validation Accuracy = 96.054%\n",
      "\n",
      "EPOCH 89 ...\n",
      "Validation Accuracy = 95.760%\n",
      "\n",
      "EPOCH 90 ...\n",
      "Validation Accuracy = 95.737%\n",
      "\n",
      "EPOCH 91 ...\n",
      "Validation Accuracy = 95.760%\n",
      "\n",
      "EPOCH 92 ...\n",
      "Validation Accuracy = 95.714%\n",
      "\n",
      "EPOCH 93 ...\n",
      "Validation Accuracy = 95.918%\n",
      "\n",
      "EPOCH 94 ...\n",
      "Validation Accuracy = 95.896%\n",
      "\n",
      "EPOCH 95 ...\n",
      "Validation Accuracy = 96.009%\n",
      "\n",
      "EPOCH 96 ...\n",
      "Validation Accuracy = 95.918%\n",
      "\n",
      "EPOCH 97 ...\n",
      "Validation Accuracy = 95.646%\n",
      "\n",
      "EPOCH 98 ...\n",
      "Validation Accuracy = 96.236%\n",
      "\n",
      "EPOCH 99 ...\n",
      "Validation Accuracy = 95.283%\n",
      "\n",
      "EPOCH 100 ...\n",
      "Validation Accuracy = 96.281%\n",
      "\n",
      "Model saved\n"
     ]
    }
   ],
   "source": [
    "with tf.Session() as sess:\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    numExamples = len(X_train)\n",
    "    \n",
    "    print(\"Training...\")\n",
    "    print()\n",
    "    for i in range(EPOCHS):\n",
    "        X_train, y_train = shuffle(X_train, y_train)\n",
    "        for offset in range(0, numExamples, BATCH_SIZE):\n",
    "            end = offset + BATCH_SIZE\n",
    "            xBatch, yBatch = X_train[offset:end], y_train[offset:end]\n",
    "            sess.run(trainingOperation, feed_dict={x: xBatch, y: yBatch, keepProb: KEEP_PROB})\n",
    "            \n",
    "        validationAccuracy = evaluate(X_valid, y_valid)\n",
    "        print(\"EPOCH {} ...\".format(i+1))\n",
    "        print(\"Validation Accuracy = {:.3f}%\".format(validationAccuracy * 100))\n",
    "        print(\"\")\n",
    "        \n",
    "    saver.save(sess, './lenet')\n",
    "    print(\"Model saved\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test the Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A validation set can be used to assess how well the model is performing. A low accuracy on the training and validation\n",
    "sets imply underfitting. A high accuracy on the training set but low accuracy on the validation set implies overfitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./lenet\n",
      "Test Accuracy = 0.943\n"
     ]
    }
   ],
   "source": [
    "with tf.Session() as sess:\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    openSession = tf.train.import_meta_graph('./lenet.meta')\n",
    "    openSession.restore(sess, \"./lenet\")\n",
    "    test_accuracy = evaluate(X_test, y_test)\n",
    "    print(\"Test Accuracy = {:.3f}\".format(test_accuracy))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "## Step 3: Test a Model on New Images\n",
    "\n",
    "To give yourself more insight into how your model is working, download at least five pictures of German traffic signs from the web and use your model to predict the traffic sign type.\n",
    "\n",
    "You may find `signnames.csv` useful as it contains mappings from the class id (integer) to the actual sign name."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load and Output the Images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f62e6fe6d8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+AkJFAY/IYHC4CckUBj8hATK/wPmC3/aXnxC+wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f630040ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nIoXiJyRSKH5CIoXiJyRSKH5CIuX/AQsN/tYjoaRmAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f627d4bda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmKLgFyLhCgDmAAAAB0lEQVSm/H89bFe1eYINpAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f634c99400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nvxAJRcEvREJR8AuRUBT8QiSU/w/uMzl1Xhg64wAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f62e79c7b8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "### Load the images and plot them here.\n",
    "### Feel free to use as many code cells as needed.\n",
    "import os\n",
    "import cv2\n",
    "xList = []\n",
    "\n",
    "for images in os.listdir(\"New_Images/\"):\n",
    "    readInImage = cv2.imread(\"New_Images/\" + images)\n",
    "    resize = cv2.resize(readInImage, (32, 32))\n",
    "    plt.figure()\n",
    "    colorConverted = cv2.cvtColor(resize, cv2.COLOR_BGR2RGB)\n",
    "    plt.imshow(colorConverted)\n",
    "    xList.append(colorConverted)\n",
    "\n",
    "xList = np.asarray(xList)\n",
    "xMyGray = np.sum(xList/3, axis = 3, keepdims = True)\n",
    "xMyNormalized = (xMyGray - 128.) / 128."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Predict the Sign Type for Each Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def sign_identifier(inputSignNum):\n",
    "    for i in signClassList:\n",
    "        if i[0] == str(inputSignNum):\n",
    "            return i[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./lenet\n",
      "--------------------------------------------------\n",
      "Image 1 Predicted Class:\n",
      "> 1 - Speed limit (30km/h)\n",
      "\n",
      "Actual Class: 1 - Speed limit (30km/h)\n",
      "--------------------------------------------------\n",
      "--------------------------------------------------\n",
      "Image 2 Predicted Class:\n",
      "> 13 - Yield\n",
      "\n",
      "Actual Class: 13 - Yield\n",
      "--------------------------------------------------\n",
      "--------------------------------------------------\n",
      "Image 3 Predicted Class:\n",
      "> 1 - Speed limit (30km/h)\n",
      "\n",
      "Actual Class: 14 - Stop\n",
      "--------------------------------------------------\n",
      "--------------------------------------------------\n",
      "Image 4 Predicted Class:\n",
      "> 1 - Speed limit (30km/h)\n",
      "\n",
      "Actual Class: 4 - Speed limit (70km/h)\n",
      "--------------------------------------------------\n",
      "--------------------------------------------------\n",
      "Image 5 Predicted Class:\n",
      "> 8 - Speed limit (120km/h)\n",
      "\n",
      "Actual Class: 8 - Speed limit (120km/h)\n",
      "--------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "### Run the predictions here and use the model to output the prediction for each image.\n",
    "### Make sure to pre-process the images with the same pre-processing pipeline used earlier.\n",
    "### Feel free to use as many code cells as needed.\n",
    "actualClass = [1,13,14,4,8]\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    openSession2 = tf.train.import_meta_graph('./lenet.meta')\n",
    "    openSession2.restore(sess, \"./lenet\")\n",
    "    pred = sess.run(logits, feed_dict={x: xMyNormalized, keepProb : 1.0})\n",
    "    pred = sess.run(tf.nn.top_k(pred, k=5, sorted = True))\n",
    "    for i in range (0,len(xMyNormalized)):\n",
    "        print(\"-\"*50)\n",
    "        print(\"Image\", i + 1, \"Predicted Class:\")\n",
    "        print(\">\", pred[1][i][0],\"-\",sign_identifier(pred[1][i][0]))\n",
    "        print()\n",
    "        print(\"Actual Class:\",actualClass[i],\"-\",sign_identifier(actualClass[i]))\n",
    "        print(\"-\"*50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Analyze Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./lenet\n",
      "--------------------------------------------------\n",
      "Percent Correct = 60.000%\n",
      "--------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "### Calculate the accuracy for these 5 new images. \n",
    "### For example, if the model predicted 1 out of 5 signs correctly, it's 20% accurate on these new images.\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    openSession3 = tf.train.import_meta_graph('./lenet.meta')\n",
    "    openSession3.restore(sess, \"./lenet\")\n",
    "    test_accuracy = evaluate(xMyNormalized, actualClass)\n",
    "    print(\"-\"*50)\n",
    "    print(\"Percent Correct = {:.3f}%\".format(test_accuracy * 100))\n",
    "    print(\"-\"*50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Output Top 5 Softmax Probabilities For Each Image Found on the Web"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Restoring parameters from ./lenet\n",
      "\n",
      "--------------------------------------------------------------------\n",
      "           Probabilities and Class Predictions:\n",
      "--------------------------------------------------------------------\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nA1xGjK8sqR+jAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f7027daf28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Speed limit (30km/h)           100.0000000000%\n",
      "Speed limit (50km/h)             0.0000000000%\n",
      "Speed limit (70km/h)             0.0000000000%\n",
      "Stop                             0.0000000000%\n",
      "Speed limit (60km/h)             0.0000000000%\n",
      "\n",
      "----------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f700daa5c0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Yield                          100.0000000000%\n",
      "Speed limit (20km/h)             0.0000000000%\n",
      "Speed limit (30km/h)             0.0000000000%\n",
      "Speed limit (50km/h)             0.0000000000%\n",
      "Speed limit (60km/h)             0.0000000000%\n",
      "\n",
      "----------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGAFJREFUeJzt3Xts1VW2B/DvEikUWq28tCpz8YHJ6HgHtSFErhOdcQYh\nJr7ujBo1RMmUGEh8gIje+IrXB0QwJF4lRY1CQOSO77fGiI8YGQvKw+HeC2qZqRCKIshDaAvr/nF+\n3FuZ31o9Z59zfqfM/n4SQrtX9++3u89ZPae/1b1/oqogovgcVukBEFFlMPmJIsXkJ4oUk58oUkx+\nokgx+YkixeQnihSTnyhSTH6iSB1eTGcRuQDAHAC9ADyuqg96X19XV6f19fUh5ylpn5Djhdq6dWvJ\nj7l//34zdthhhf889/7K0zueN4+7d+9Obd+1a5fZp6Ojw4wNGTLEjHlj7NWrV2q797icfPLJZsyb\nq969e5uxEN651q1bl9q+Z88edHR05PUED05+EekF4D8A/BZAK4BPReRlVf2L1ae+vh7z588v+FzW\nA+g9+bwHwjped6zzeQ/SM888EzSOffv2mbE9e/aYsX79+qW2ez8wvFjfvn3NmDf+lStXprYvW7bM\n7LNp0yYzdvXVV5ux6upqMzZgwIDU9oULF5p9XnvtNTMW+gPKe45Yz6vOzk6zz9ixY1PbV6xYYfY5\nWDFv+0cCWK+qX6lqO4DFAC4q4nhElKFikv84AH/r8nlr0kZEh4Bikj/tvcrfvbcRkUYRaRaR5m3b\nthVxOiIqpWKSvxXA0C6fHw9g48FfpKpNqtqgqg11dXVFnI6ISqmY5P8UwHAROUFEqgBcAeDl0gyL\niMot+Gq/qnaKyGQAbyFX6ntSVb/Io19qu3d13upz+OFFVSpTeRUE68q3V8FobW01YwsWLDBjNTU1\nZiykStDe3m728XjzEXIF2+M9nrNmzTJjU6dONWPbt29PbT/zzDPNPv379zdjb7/9thm78MILzVgI\n73F+/PHHU9svvfTSvI9fVPao6usAXi/mGERUGfwLP6JIMfmJIsXkJ4oUk58oUkx+okhJlvv2n3rq\nqWotqPBKQ1bJoxylvj59+pgxa+xvvfWW2Wf16tVmzFtQ45V5vLnyFoMcykJXHg4cODC1/dprrzX7\neGXn8ePHm7GNG//ub9z+z+mnn27GQhb2WM+Ps88+G8uXL8+rzspXfqJIMfmJIsXkJ4oUk58oUkx+\nokiV/nJ5IO8Kdqmv6nvHW7RokRl7/vnnU9tbWlrMPt6V6ND94LxtvP5RhVSDAPvq/KOPPmr2mThx\nohmzFtQAwC233GLGSr3QyYoVch6+8hNFislPFCkmP1GkmPxEkWLyE0WKyU8UqUwX9px22mm6ePHi\ngvuFlPq8RRGvvPKKGbPKeQCwYcOG1Hbv7jqh8xu6P16pF/aEllmtcXglTG/s3nx4Y7TmyisPeuO4\n/vrrzZi3KGzSpElm7IgjjjBjFut7HjlyJJqbm7mwh4hsTH6iSDH5iSLF5CeKFJOfKFJMfqJIFbVc\nTkRaAOwAsA9Ap6o2hB7L28/O8u2335qx999/34w99dRTZqytrc2MebfQsoSW3kJWgQF2Cagce/uF\nlDG9xzm0nBdyPm+1paepqcmM3XrrrWbMW0W4Y8eO1Pb7778//4EFKMVa2fNU1c5CIuqR+LafKFLF\nJr8CeFtElotIYykGRETZKPZt/2hV3SgiQwC8IyL/paofdP2C5IdCIwDU19cXeToiKpWiXvlVdWPy\nfxuAFwCMTPmaJlVtUNWGo446qpjTEVEJBSe/iPQXkdoDHwP4HYA1pRoYEZVXMW/7jwbwQlKSOhzA\nIlV9M/RgVVVVZmznzp2p7UuXLjX7zJ8/34xt27bNjHkrs3788cfU9nLcNswriYWUAcuxunD37t1m\nrF+/fgWfy1tp5yl1iTD08Zw5c6YZ827Xde6556a2z5492+wzbdq0vMdlCX7WqupXAH5Z9AiIqCJY\n6iOKFJOfKFJMfqJIMfmJIsXkJ4pU5vfqs0pY3go9q6T37LPPmn2slVLFsEpAXqkpZLUiEL5Jp9XP\nO543Ru9cXlk0hHeukPvWebz58Fb8hT6en332mRlbtmxZavttt91W8nF0xVd+okgx+YkixeQnihST\nnyhSTH6iSGV6tb+zsxPfffddauzDDz80+y1YsCC13Vrw0x3v6rB36y3rarS3ICX0Krt3WyvvmLt2\n7Upt9xbaeOP3rnx7serq6tT2vXv3mn28x6W9vd2MTZ48ueBjevPrVYq8PfxC90m0qiYh+x0WsoCL\nr/xEkWLyE0WKyU8UKSY/UaSY/ESRYvITRSrTUt/OnTvxySefpMasxQ0A8P3336e2h5aotm/fbsa8\nHYZDFlOcf/75Zuyss84yY17Zy/verH0GH3rooaDj3XzzzWbMK5dZJSdvDr0xeo+1t/+jdT5rngCg\nf//+ZmzKlClm7N577zVj3lxZJUKvpGuVuQt5jvKVnyhSTH6iSDH5iSLF5CeKFJOfKFJMfqJIdVvq\nE5EnAVwIoE1Vf5G0DQDwLIBhAFoA/EFV0+txB7FWzTU3N5t9rFKOt9LrlltuMWMdHR1mzCuVWKuv\nZsyYYfYZO3asGVu/fr0Zq6urM2OzZs0yYyElNut2Ud7xAGDu3Llm7LLLLkttHzx4sNnHKzl6qwFD\nSrCPPfaYGautrTVj1113nRmzVjIC/nPOKmN6z++XXnoptd27Fd3B8nnlfwrABQe1TQfwrqoOB/Bu\n8jkRHUK6TX5V/QDA1oOaLwLwdPLx0wAuLvG4iKjMQn/nP1pVNwFA8v+Q0g2JiLJQ9gt+ItIoIs0i\n0mztMkNE2QtN/s0iUg8Ayf9t1heqapOqNqhqg/c300SUrdDkfxnA+OTj8QDSLz0SUY+VT6nvGQDn\nAhgkIq0A7gLwIIAlIjIBwF8B/D7fE1qlI6/M45VJLF5pyFs9du2115qxn/3sZ6nt3uqrqVOnmjFv\n5Z7Xz9tk1Jpfb+XbOeecY8a8ufdWxq1cuTK1fcyYMWYfr6zo3RrMK5mecsopqe3eHG7devD17f/n\nlRW9Y3qs54+XE+vWrUtt37NnT97n7Tb5VfVKI/SbvM9CRD0O/8KPKFJMfqJIMfmJIsXkJ4oUk58o\nUplu4An45YtCnXjiiWbMK7t4ZaPW1lYz9uabb6a2eyUqrzQUem+3kHvaeeVIbz68cp7nhx9+KKgd\n8J8b3uO5fPlyM3byySentnuPi7fBayhv/kNWYlqPWSH5xVd+okgx+YkixeQnihSTnyhSTH6iSDH5\niSKVeanPErIJ46hRo4KOt3v3bjO2dOlSM2aVa7wyjleW81baebxNGmtqalLbvXvdbdmyxYyF3GPO\n6+dtcumVTL0SlrcKz+Ldg9DjjdFbAek9R6zvzSvBWqv3eK8+IuoWk58oUkx+okgx+YkixeQnilTm\nV/utq6UhtzOaN2+e2Wf6dPsmQrfeeqsZ864qP/LII6ntIXsMAv6VY8+AAQPMmLWwx1sY413R9yoS\n3lxZC4K8q96hvMVCFm8b+TfeeMOMffnllwWfCwhb0OaNsW/fvqnthcwvX/mJIsXkJ4oUk58oUkx+\nokgx+YkixeQnilQ+t+t6EsCFANpU9RdJ290A/gjgwIqQ21X19XxOaJW3vAUwVp+6ujqzzwMPPGDG\nRo8ebcbGjRtnxqyFRB9//LHZx1toEbqfYcgec7W1tWYsZFEV4D9m1u3SvPKmt1DIK2F5/aw5njt3\nrtnH440jNGbNifeYebejy1c+z76nAFyQ0v6wqo5I/uWV+ETUc3Sb/Kr6AYDC10wSUY9WzO/8k0Vk\nlYg8KSJHlWxERJSJ0OR/DMBJAEYA2ARglvWFItIoIs0i0uz9uSIRZSso+VV1s6ruU9X9AOYBGOl8\nbZOqNqhqQ//+/UPHSUQlFpT8IlLf5dNLAKwpzXCIKCv5lPqeAXAugEEi0grgLgDnisgIAAqgBcDE\nYgfi7TFnlaKsfcwAoF+/fmZs5cqVZmzMmDFmzFr9FlrOC10N6K3Q88qfpT6XV7azSlHefISu+PMe\na2v85Vhd6M2VxxrL119/bfY59thjg87VVbfJr6pXpjQ/UfSZiaii+Bd+RJFi8hNFislPFCkmP1Gk\nmPxEkcp0A08RcTeLLJR3y6UlS5aYsc2bN5sxr3z14YcfFtzHc/zxx5sxryTmlfOslXZeacsrVXrl\nK2+M55xzTmq7V970nhteP++2Vtb35pWWQ0t23nx4zxEr9tFHH5l9Lr/88vwHZuArP1GkmPxEkWLy\nE0WKyU8UKSY/UaSY/ESRyrTUt2PHDrNcFrKJpHVfOsDfiLOpqcmMeSUZ63yhK8TOO+88Mxa6qaY1\nfm+Ty0cffdSMefc1vOmmm8yYdT5v009P6AaepeaV80IfM8uGDRvMmFWq5L36iKhbTH6iSDH5iSLF\n5CeKFJOfKFKZX+1/7733UmNHHHGE2c+6gu1dXfWOd8UVV5gx7zZI1pXU0IU9w4YNM2Oh25xbY6yq\nqjL7eHshzpkzx4x5C6taWlpS21988UWzj7d4xxu/95hZV+e979lbYJTl1X6vMmJVOAp5LvKVnyhS\nTH6iSDH5iSLF5CeKFJOfKFJMfqJISXelAREZCmA+gGMA7AfQpKpzRGQAgGcBDEPull1/UNXvvWNV\nVVXp4MGDU2Pe4gxrjN7YvXJNW1ubGfNKhNZiCq9E5e0H543R25fO+76tElZtba3ZJ3RhTEg/bz68\nffVKvXjHK6N58+uV87xYyB5+3nxYZda5c+fim2++yWt1Tz6v/J0ApqjqzwGMAjBJRE4FMB3Au6o6\nHMC7yedEdIjoNvlVdZOqrkg+3gFgLYDjAFwE4Onky54GcHG5BklEpVfQ7/wiMgzAGQCWAThaVTcB\nuR8QAIaUenBEVD55/3mviNQAeA7Ajar6Q76bBohII4BGwP8dhoiyldcrv4j0Ri7xF6rq80nzZhGp\nT+L1AFKvoqlqk6o2qGqD93fRRJStbrNRci/xTwBYq6qzu4ReBjA++Xg8gJdKPzwiKpd83vaPBnAN\ngNUi8nnSdjuABwEsEZEJAP4K4PfdHaizsxNbt25NjR155JFmP6sU4pWNvJh3rpB3J15Zzttn0FuN\nFlrqq66uTm0vxz53oXsXZnU8j/f88MqA3vPDm2OvDGiViqdNm1bwuQpZ1ddt8qvqRwCsR+U3eZ+J\niHoU/hJOFCkmP1GkmPxEkWLyE0WKyU8UqUw38Kyvr8ekSZNSY4sWLTL7bd68ueBzZVk2ClmRCPgl\nR4+3GjBks9Mshd6uq9THLMfzwzumt/LTuyWa5c4770xtf/XVV/M+Bl/5iSLF5CeKFJOfKFJMfqJI\nMfmJIsXkJ4pUpqU+wC45effPe/jhh1PbQ+5lBmRbBgwt/3gbn3gry6xYTyn1HQqrC73nlbdK07uf\n4IQJE8xY3759U9svv/xys08p9sbgKz9RpJj8RJFi8hNFislPFCkmP1GkMr3a397ejg0bNqTGhg8f\nbvabMmVKavucOXPMPt6CGu/qsHeV3Tqmd9XeE3rLKO+Kubfop9S8ubKqC6GPixfz5sOa49AFRt78\nelf0vUVcra2tqe3Dhg3Le1wh+MpPFCkmP1GkmPxEkWLyE0WKyU8UKSY/UaSku9v7iMhQAPMBHANg\nP4AmVZ0jIncD+COALcmX3q6qr3vHOuOMM3Tp0qWpsVmzZpn9rPJKv379zD733XefGQstlXm3eArh\nncsro3llL6uEFbqgJrQ0F1LqK8eiH2scu3btMvsMGjTIjN14441B49i+fbsZu//++4OOmaahoQHN\nzc15rXTKp9jZCWCKqq4QkVoAy0XknST2sKo+FDpQIqqcfO7VtwnApuTjHSKyFsBx5R4YEZVXQb/z\ni8gwAGcAWJY0TRaRVSLypIgcVeKxEVEZ5Z38IlID4DkAN6rqDwAeA3ASgBHIvTNI/aVdRBpFpFlE\nmr/77rsSDJmISiGv5BeR3sgl/kJVfR4AVHWzqu5T1f0A5gEYmdZXVZtUtUFVGwYOHFiqcRNRkbpN\nfsld0n0CwFpVnd2lvb7Ll10CYE3ph0dE5ZLP1f7RAK4BsFpEPk/abgdwpYiMAKAAWgBM7O5AImLu\nc3bbbbeZ/awy4I8//mj2GTp0qBlraWkxY6UWur+cV1YMuV1XKG/1mzf/1h5zpdh7rhDW+Wpqasw+\nXjnP26fP2yexlOW8Usnnav9HANKewW5Nn4h6Nv6FH1GkmPxEkWLyE0WKyU8UKSY/UaQy3cBTRMzS\nkVcmmTZtWmr7zJkzzT7jx483Y/fcc48Z80pzpb4Vllf2Ct3cM2Q1nccrOXpjtL630JWRIZt0Avb3\nfdVVV5l9vMfFK7NOmjTJjPVEfOUnihSTnyhSTH6iSDH5iSLF5CeKFJOfKFKZlvo8Iau9vJWADz1k\nby04ffp0MzZjxgwz5m2qGSJ0c0yvX6nLkSHn8oSucgztd8wxx6S2n3TSSWaf6upqM9bY2GjGQsdY\nKXzlJ4oUk58oUkx+okgx+YkixeQnihSTnyhSPabU57HKTR0dHWaf9vZ2M1ZbW2vGvI0W77jjjtT2\n0BKPVyoLLbGF3O/OG79X3gxZoRe6utAbx969e83Yddddl9ruzdPEifZetIdaOc/DV36iSDH5iSLF\n5CeKFJOfKFJMfqJISXdXX0WkL4APAPRBrjrwJ1W9S0ROALAYwAAAKwBco6r2JXYAZ511ln788ccF\nD9K6uh165Xj27NlmzNujzeItIvLG2KdPHzPmXdH3KhkWb5+7Uu+P5/GulntX7b2Yt1DLGuMNN9xg\n9vEW9vR0DQ0NaG5uzqskkc8r/14Av1bVXyJ3O+4LRGQUgBkAHlbV4QC+BzAhdMBElL1uk19zdiaf\n9k7+KYBfA/hT0v40gIvLMkIiKou8fucXkV7JHXrbALwD4EsA21T1wPvFVgDHlWeIRFQOeSW/qu5T\n1REAjgcwEsDP074sra+INIpIs4g0b9myJXykRFRSBV3tV9VtAJYCGAWgTkQOXA06HsBGo0+Tqjao\nasPgwYOLGSsRlVC3yS8ig0WkLvm4GsD5ANYCeA/AvyZfNh7AS+UaJBGVXj4Le+oBPC0ivZD7YbFE\nVV8Vkb8AWCwi/w7gMwBP5HNCq/QScusnj7dn3dSpU83YnDlzzJi1kOimm24y+4SMvRy8cp5XfvMe\nF2+OrZJpyMIjwB+jd8xRo0alth/K5bxS6Tb5VXUVgDNS2r9C7vd/IjoE9YyXJSLKHJOfKFJMfqJI\nMfmJIsXkJ4pUt6v6SnoykS0ANiSfDgLwbWYnt3EcP8Vx/NShNo5/UtW8/pou0+T/yYlFmlW1oSIn\n5zg4Do6Db/uJYsXkJ4pUJZO/qYLn7orj+CmO46f+YcdRsd/5iaiy+LafKFIVSX4RuUBE/ltE1ouI\nvfti+cfRIiKrReRzEWnO8LxPikibiKzp0jZARN4RkXXJ/0dVaBx3i8g3yZx8LiLjMhjHUBF5T0TW\nisgXInJD0p7pnDjjyHRORKSviPxZRFYm47gnaT9BRJYl8/GsiFQVdSJVzfQfgF7IbQN2IoAqACsB\nnJr1OJKxtAAYVIHz/grAmQDWdGmbCWB68vF0ADMqNI67AUzNeD7qAZyZfFwL4H8AnJr1nDjjyHRO\nAAiAmuTj3gCWIbeBzhIAVyTtcwFcX8x5KvHKPxLAelX9SnNbfS8GcFEFxlExqvoBgK0HNV+E3Eao\nQEYbohrjyJyqblLVFcnHO5DbLOY4ZDwnzjgypTll3zS3Esl/HIC/dfm8kpt/KoC3RWS5iDRWaAwH\nHK2qm4DckxDAkAqOZbKIrEp+LSj7rx9dicgw5PaPWIYKzslB4wAynpMsNs2tRPKnbclSqZLDaFU9\nE8BYAJNE5FcVGkdP8hiAk5C7R8MmALOyOrGI1AB4DsCNqvpDVufNYxyZz4kWsWluviqR/K0Ahnb5\n3Nz8s9xUdWPyfxuAF1DZnYk2i0g9ACT/t1ViEKq6OXni7QcwDxnNiYj0Ri7hFqrq80lz5nOSNo5K\nzUly7oI3zc1XJZL/UwDDkyuXVQCuAPBy1oMQkf4iUnvgYwC/A7DG71VWLyO3ESpQwQ1RDyRb4hJk\nMCeS26DvCQBrVbXrvdQynRNrHFnPSWab5mZ1BfOgq5njkLuS+iWAf6vQGE5ErtKwEsAXWY4DwDPI\nvX3sQO6d0AQAAwG8C2Bd8v+ACo1jAYDVAFYhl3z1GYzjX5B7C7sKwOfJv3FZz4kzjkznBMA/I7cp\n7irkftDc2eU5+2cA6wH8J4A+xZyHf+FHFCn+hR9RpJj8RJFi8hNFislPFCkmP1GkmPxEkWLyE0WK\nyU8Uqf8FnGkvxiv6McYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f700b9bc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Speed limit (30km/h)            99.9953508377%\n",
      "Stop                             0.0033806267%\n",
      "Yield                            0.0012675282%\n",
      "Speed limit (50km/h)             0.0000002531%\n",
      "Keep left                        0.0000000453%\n",
      "\n",
      "----------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f6d41d0898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Speed limit (30km/h)            99.9999403954%\n",
      "Speed limit (60km/h)             0.0000351977%\n",
      "Stop                             0.0000248258%\n",
      "Speed limit (70km/h)             0.0000027374%\n",
      "Turn right ahead                 0.0000000031%\n",
      "\n",
      "----------------------------------------------------------------------\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1f6fe1124a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Speed limit (120km/h)           73.9777505398%\n",
      "Speed limit (20km/h)            22.8198796511%\n",
      "Speed limit (30km/h)             2.8353558853%\n",
      "Go straight or right             0.2367284149%\n",
      "Keep right                       0.0944264291%\n",
      "\n",
      "----------------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "### Print out the top five softmax probabilities for the predictions on the German traffic sign images found on the web. \n",
    "### Feel free to use as many code cells as needed.\n",
    "\n",
    "with tf.Session() as sess:\n",
    "    sess.run(tf.global_variables_initializer())\n",
    "    openSession4 = tf.train.import_meta_graph('./lenet.meta')\n",
    "    openSession4.restore(sess, \"./lenet\")\n",
    "    softmax = tf.nn.softmax(logits)\n",
    "    topK = tf.nn.top_k(softmax, k=5)\n",
    "    pred = sess.run(softmax, feed_dict={x: xMyNormalized, keepProb: 1.0})\n",
    "    pred = sess.run(topK, feed_dict={x: xMyNormalized, keepProb: 1.0})\n",
    "    print()\n",
    "    print(\"-\" * 68)\n",
    "    print(\" \" * 10, \"Probabilities and Class Predictions:\")\n",
    "    print(\"-\" * 68)\n",
    "    print()\n",
    "    \n",
    "    for i in range (0,len(xMyNormalized)):\n",
    "        plt.imshow(xMyNormalized[i].squeeze(), cmap='gray')\n",
    "        plt.show()\n",
    "        for j in range (0,len(xMyNormalized)):\n",
    "            print('{:<30} {:>14.10f}%'.format(sign_identifier(pred[1][i][j]), pred[0][i][j]*100))\n",
    "        print()\n",
    "        print(\"-\" * 70)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Project Writeup\n",
    "\n",
    "Once you have completed the code implementation, document your results in a project writeup using this [template](https://github.com/udacity/CarND-Traffic-Sign-Classifier-Project/blob/master/writeup_template.md) as a guide. The writeup can be in a markdown or pdf file. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "> **Note**: Once you have completed all of the code implementations and successfully answered each question above, you may finalize your work by exporting the iPython Notebook as an HTML document. You can do this by using the menu above and navigating to  \\n\",\n",
    "    \"**File -> Download as -> HTML (.html)**. Include the finished document along with this notebook as your submission."
   ]
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "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": "3.5.2"
  },
  "widgets": {
   "state": {},
   "version": "1.1.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}