one_shot.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# One-shot learning (N-way) using Omniglot data set and Siamese Neural Networks\n",
    "\n",
    "This notebook shows an implementation version of the paper [koch et al, Siamese Neural Networks for One-shot Image Recognition](http://www.cs.cmu.edu/~rsalakhu/papers/oneshot1.pdf).\n",
    "\n",
    "The model has been created with [keras](https://keras.io/).\n",
    "\n",
    "**NOTE: Bear in mind this is a work in progress repository and there is plenty of room for improvements in the components as well as for the one-shot experiments.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## One-shot learning (N-way)\n",
    "\n",
    "Humans learn new concepts with very little supervision – e.g. a child can generalize the concept\n",
    "of “giraffe” from a single picture in a book.\n",
    "\n",
    "Deep learning has made major advances in areas such as speech, vision and language, but is notorious for requiring large datasets.\n",
    "\n",
    "The process of learning good features for machine learning applications can be very computationally expensive and may prove difficult in cases where little data is available.\n",
    "\n",
    "One-shot learning is a clasification problem in machine learning, that aims to learn information about classes from one, or only a few, training examples.\n",
    "\n",
    "A one-shot task aims to classify a test example X, into one of the N unique classes (N-way) of a support set S, which is a small subset of the all available C classes.\n",
    "\n",
    "The model is given the support set and a test example that has to classify.\n",
    "\n",
    "Since exactly one example in the support set has the right class of the test example, the aim is to correctly predict which support example is the same as the test example."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Omniglot data set\n",
    "\n",
    "The data set contains 50 alphabets total, splitted into a background set of 30 alphabets and an evaluation set of 20 alphabets. There is no overlap between the training and evaluation sets.\n",
    "\n",
    "Each alphabet has a number of characters from 14 to 55 different characters drawn by 20 different subjects, resulting in 20 105x105 images for each character with a total of 1623 character classes.\n",
    "\n",
    "The background set is only used to learn general knowledge about characters (e.g., hyperparameter inference or feature learning). One-shot learning results are reported using alphabets from the evaluation set.\n",
    "\n",
    "![alt text](./images/omniglot_general.png \"Omniglot Data Set\")\n",
    "\n",
    "Following are some of the alphabets represented in the data set:\n",
    "\n",
    "![alt text](./images/omniglot_alphabets.png \"Omniglot Alphabets\")\n",
    "\n",
    "An example of a one-shot 20-way task in this dataset can be seen in the following figure:\n",
    "    \n",
    "![alt text](./images/omniglot_20_way.png \"Omniglot 20-way task\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Omniglot Service\n",
    "\n",
    "The data set is retrieved using OmniglotService class.\n",
    "\n",
    "It download and unzip the data set to the path specified when creating the class (./data by default) within the train_set and test_set folders respectively."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://github.com/brendenlake/omniglot/raw/master/python/images_background.zip\n",
      "Unziping data/images_background.zip to data\n",
      "Rename data/images_background to data/train_set\n",
      "Downloading data from https://github.com/brendenlake/omniglot/raw/master/python/images_evaluation.zip\n",
      "Unziping data/images_evaluation.zip to data\n",
      "Rename data/images_evaluation to data/test_set\n"
     ]
    }
   ],
   "source": [
    "from omniglot_service import OmniglotService\n",
    "\n",
    "data_service = OmniglotService()\n",
    "data_service.get_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "You can also retrieve the data by set type (train/test) and specify the folder name where will be unzip."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://github.com/brendenlake/omniglot/raw/master/python/images_evaluation.zip\n",
      "Unziping data/images_evaluation.zip to data\n",
      "Rename data/images_evaluation to data/test_alphabets\n"
     ]
    }
   ],
   "source": [
    "data_service.get_data_type('test_alphabets', data_type='test')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data folder content after getting the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[34mtest_alphabets\u001b[m\u001b[m \u001b[34mtest_set\u001b[m\u001b[m       \u001b[34mtrain_set\u001b[m\u001b[m\n"
     ]
    }
   ],
   "source": [
    "!ls ./data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below there is a summary of the folders content:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Set:\n",
      "Number of alphabets =       30\n",
      "First five alphabets = ['Alphabet_of_the_Magi', 'Anglo-Saxon_Futhorc', 'Arcadian', 'Armenian', 'Asomtavruli_(Georgian)']\n",
      "First five classes of first alphabet = ['character01', 'character02', 'character03', 'character04', 'character05']\n",
      "First five images of first character of first alphabet = ['0709_01.png', '0709_02.png', '0709_03.png', '0709_04.png', '0709_05.png']\n",
      "\n",
      "Test Set:\n",
      "Number of alphabets  =       20\n",
      "First five alphabets = ['Angelic', 'Atemayar_Qelisayer', 'Atlantean', 'Aurek-Besh', 'Avesta']\n",
      "First five classes of first alphabet = ['character01', 'character02', 'character03', 'character04', 'character05']\n",
      "First five images of first character of first alphabet = ['0965_01.png', '0965_02.png', '0965_03.png', '0965_04.png', '0965_05.png']\n"
     ]
    }
   ],
   "source": [
    "print(\"Train Set:\")\n",
    "number_alphabets = !ls ./data/train_set | wc -l\n",
    "print(f'Number of alphabets = {number_alphabets[0]}')\n",
    "alphabets = !ls ./data/train_set | head -5\n",
    "print(f'First five alphabets = {alphabets}')\n",
    "classes = !ls ./data/train_set/Alphabet_of_the_Magi | head -5\n",
    "print(f'First five classes of first alphabet = {classes}')\n",
    "images = !ls ./data/train_set/Alphabet_of_the_Magi/character01 | head -5\n",
    "print(f'First five images of first character of first alphabet = {images}')\n",
    "print()\n",
    "print(\"Test Set:\")\n",
    "number_alphabets = !ls ./data/test_set | wc -l\n",
    "print(f'Number of alphabets  = {number_alphabets[0]}')\n",
    "alphabets = !ls ./data/test_set | head -5\n",
    "print(f'First five alphabets = {alphabets}')\n",
    "classes = !ls ./data/test_set/Angelic | head -5\n",
    "print(f'First five classes of first alphabet = {classes}')\n",
    "images = !ls ./data/test_set/Angelic/character01 | head -5\n",
    "print(f'First five images of first character of first alphabet = {images}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Omniglot Loader\n",
    "\n",
    "The data set is loaded into memory from the specified data set types folders using OmniglotLoader class.\n",
    "\n",
    "If any of the folders doesn't exist the loader uses the OmniglotService to retrieve de data set.\n",
    "\n",
    "Each data set type is stored in a tuple respectively with the first element being the data and the second the classes in the format 'alphabet/character' and saved to a pickle file to load it more quickly in subsequents loads."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading data from data/train_set\n",
      "Saving data to data/train.pickle\n",
      "Loading data from data/test_set\n",
      "Saving data to data/test.pickle\n"
     ]
    }
   ],
   "source": [
    "from omniglot_loader import OmniglotLoader\n",
    "\n",
    "omniglot_loader = OmniglotLoader(path='./data', train_folder='train_set', test_folder='test_set')\n",
    "train_set, test_set = omniglot_loader.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading data from data/train.pickle\n",
      "Loading data from data/test.pickle\n"
     ]
    }
   ],
   "source": [
    "train_set, test_set = omniglot_loader.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below there is a summary of the tuple's content:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Set:\n",
      "Number of classes = 964\n",
      "Images per class  = 20\n",
      "Image shape (H,W,C) = 105, 105, 1\n",
      "First five classes = ['Alphabet_of_the_Magi/character01' 'Alphabet_of_the_Magi/character02'\n",
      " 'Alphabet_of_the_Magi/character03' 'Alphabet_of_the_Magi/character04'\n",
      " 'Alphabet_of_the_Magi/character05']\n",
      "\n",
      "Test Set:\n",
      "Number of classes = 659\n",
      "Images per class  = 20\n",
      "Image shape (H,W,C) = 105, 105, 1\n",
      "First five classes = ['Angelic/character01' 'Angelic/character02' 'Angelic/character03'\n",
      " 'Angelic/character04' 'Angelic/character05']\n",
      "\n",
      "Number of classes in both data sets = 0\n"
     ]
    }
   ],
   "source": [
    "print(\"Train Set:\")\n",
    "classes, images_per_class, height, width, channels = train_set[0].shape\n",
    "print(f'Number of classes = {classes}')\n",
    "print(f'Images per class  = {images_per_class}')\n",
    "print(f'Image shape (H,W,C) = {height}, {width}, {channels}')\n",
    "print(f'First five classes = {train_set[1][0:5]}')\n",
    "print()\n",
    "print(\"Test Set:\")\n",
    "classes, images_per_class, height, width, channels = test_set[0].shape\n",
    "print(f'Number of classes = {classes}')\n",
    "print(f'Images per class  = {images_per_class}')\n",
    "print(f'Image shape (H,W,C) = {height}, {width}, {channels}')\n",
    "print(f'First five classes = {test_set[1][0:5]}')\n",
    "print()\n",
    "common_classes = set(train_set[1])&set(test_set[1])\n",
    "print(f'Number of classes in both data sets = {len(common_classes)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Following is a random sample of each data set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAD0CAYAAACYaSJ/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGdVJREFUeJzt3Xm4HFWZx/HfD8ISEyRCEB0IRGHGBVFHUQSXQQUBEVEBRUUFiYI8zuAjCA4ogqKgouD6wIwLBFwQdBgBUWQ0RpagMIAsig6YBSSYhZCAAQTe+eOchkqnu2/3vbdvL+f7eZ77PLe6Ttc51ff2qbfeOlXHESEAAIBSrdPrBgAAAPQSwRAAACgawRAAACgawRAAACgawRAAACgawRAAACgawdA4sb2u7fttb9XrtoyF7StsH9TrdgAAOmd7nu0De92OQVNsMJQDl9rPY7ZXV5bf2en2IuLRiJgaEQtH2Z73277N9irbi21fbHvKaLYFACUa7369st0RAwzbH7D9x1zXYtsX2Z482joxsSb1ugG9EhFTa7/bni9pVkRc3qy87UkR8Ug32mL7tZJOlLRHRNxoe1NJe3ejLgAYVp326+PF9u6SPqbUh9+U+/B9ul0vxk+xmaGR2D7J9nm2v2d7laQDbe+UzxBW2L7b9pdtr5fLT7Idtmfm5XPz+ktztudq289oUt1LJF0ZETdKUkQsi4izIuKBvK032r4hb2eh7Y9X2rltrvcg23faXm77fbZ3tH1TbuuXKuVn2Z5r++u277P9e9uvbvE5zLL9B9v35n2ZMcaPFgB6Ig9n+LjtO2wvtf0d29Pyuim2v5/70BW2r7H9FNtfUOqjv5GzPl9osOmXSPp1RNwkPd6HfysiVudtv9n2jbZX2l5g+9hKm55t+xHbh9i+y/Yy2+/Nx5ubc1u+WCl/mO1f2D4zb+9W269qsc+H5qsOy21fYnuL8fk0hwvBUGtvlvRdSRtLOk/SI5KOkDRd0ssl7SHp0Bbvf4ekj0vaRNJCSZ9qUm6epL1sf8L2zrY3qFt/v6QDczv2lnSE7TfUldlB0ja53JclfVTSayQ9TymQe3ml7M6S/pD341OS/qvWIVTZ3k/SR5TOcDaTdE3+PGrrL7V9VIv9B4B+8hFJr5P0CklbSvq7pNPyullKV0u2UOobPyjp4Yg4UtJvlbJMU/NyvXmS3mj7+BzErF+3fqXS8WCa0nHlKNt7VNavK+n5kp4p6WBJX5F0lKR/ya8fbHvHSvlXSbpR0qaSTpF0oe0n1zfK9gGSPqR03Nhc0vWSzq2s/7ntDzX5rIpCMNTaFRFxUUQ8FhGrI+K3EXFNRDwSEXdI+g+lf9ZmLoiIayPi75K+I+mFjQpFxBxJ+ymdXVwqaantz9teJ6//RUTcnNtxo6TvN6j3UxHxUET8RNLDks6NiCURcaekKyT9c6Xs3ZK+EhF/j4jvSrpD0p4NmnaopM9ExG35EuFJkl5aO7OIiD0j4tQW+w8A/eRQSR+NiL9ExINKwxPeZttKgdFmkrbJffxva9n5keRLcQdI2lHSz5T68M9W+vD/iYhbch/+v5J+oLX78E/mPvzHeXl2zjAtlHSV1uzDF0XE13MfPlvSnZJ2b7K/J0XEH/Nx6ERJr7C9eW7XbhFxejv7OOwIhlpbVF3I6cxL8uC4lZI+qXQG0cziyu9/kzS1WcGIuCQi3iDpKZLeIul9SmcIymcac2wvsX2f0hnM9Lr331NZXC2pfrla952x5gy9CyT9Q4NmbS3pazlNu0LSUkmPKZ1RAcDAyAHPDEk/qfRp1ysdBzeV9E1Jv5J0QR5y8Bnb67a7/Yj4cUTspZT92V/SByS9K9f9ctu/qvThB2nNPvzRiFhWWR6xD6+rvlUffkZlf5coXeGgD69DMNRa1C2fKelmSdtGxJMlHS/J41phOnP4uaQ5Spe4pJQJ+qGkGRGxsaRvjLHe+i/CVpL+0qDcIkmHRMS0ys/kiLhmDHUDwITLJ4B3SXpNXZ+2YUQszVmZ4yPi2UqXofZXyvZIax8LWtXzWET8TNJcPdGH/0BpqEWtDz9LE9eHH9SgD79uDHUPJYKhzmwk6T5JD9h+jlqPF2pbHlz31jxYz7ZfJumVSteha/Uuj4gH87oDmm6sPU+3/UGnQd8HKI01+mmDcmdIOi7vq2xPy+OIAGAQnSHplNqNILafanvv/Puutp+bL22tVMqgPJrfd4/SeJ6GbO9ne//cR9r2zkrjSufljNRUSctyH76zUqA1FjPyQOpJTrf8byXpsib7+zHbz8rtfIrtfcdY91AiGOrMkZLeI2mVUpbovHHa7gpJh0n6P6Uv4dlKY3Vq2/+ApJOd7mo7VuksYyyukrSdpOWSTpC0b0TcW18oIs6X9EVJ5+fLgr9T5bq07ctsHz3GtgDARPmcpMsl/SL3p1dJelFet4Wk/1bq32+W9BM90deeJundTnfVfq7Bdu+VdLik25X68G9JOjEifpgzUodJOjXXebSk88e4H3OVxhAtl3ScpDdHxH31hSLie5K+KulHuQ+/QdJutfX5rrQPj7EtQ8FrDh3BsLM9S9KBEbFLr9sCAOiM7cMk7RcRu/a6LcOEzBAAACgawRAAACgal8kAAEDRyAwBAICiFRUMOc0r0/T2yC7Ud4vtXUbxvrC9bReaBAB9zfbM3Ad2dSJx22fZPqmbdbTRhhNsnztySXTbwAVDtg9wmkDvAdt/zb8fnp/l0FKeV+aOLrVrrS9WRGyXp9qolrvM9uu60YbxZnu+7THdsWD7VNt/cppk9g+23123PvLf8v78842xtRpAL+V+42Hb0+tev8GVyayHndM8k1f1uh3tcJro+4oxbmN92xfkv3/UJwLyM5jOzsftv9o+YSz1jbeBCoZsHynpS5I+L+lpShPPHab0cKv6ifHGs95xOUOxPUXSi5Ue+d513T6zGqFu54eXPaA0SeDGSs9o+lJ+6FjVC3KgOjUiZk10WwGMuz9Lenttwfb2kib3rjk98XqlZxVNiB7397W6r1CaLHxxg2KnSXqSpJmSXirpXbYPnpAGtmFggiHbGyvNBXZ4RFwQEasiuT4i3hkRDznN3zWr8p41ot3q5Sfbe9m+3vZK24uqUWolTXuI7YWSfpFfP99pXrL7bM+1vV1+/f2S3inp6JzduCi/Xp9Zea2kK3Nb17V9rO3bc9bkOuenoma75ozKvba/Vst82d4mPyhrme2ltr/jyozzuc5jbP9O6UnZk2x/tFLPrbbfXPfZvs/27yvrX2T7HKWnml6U9+noXPZltq9ymuvmxmr0nz//T9u+UmkutmdGxCci4g/5EfXXSPq1pJ06+uMDGDTnSKpmgd8jaXZtoVX/W8/2wZX+6Q7bh1bW3ez8BOm8vF7uF1+Ylxv22Q3qeIrti53mDrs3/75lZf0c2yflvu9+2xfZ3jT3vytt/9ZrZ7weD4Zsb+c0Q/xy2/fYPrZSbn3bs/P+3WJ7h0q9TfvufHy70vZptpdLOqGN48MM2z/K+7nM9ledZhg4Q9JOed9W5LIbOGX2F+Y2n2F7cl63i9P8bcfYXizp2xHxcEScHhFX6Iknd1ftLelzEfG3iJivNBfcexv9PXoiIgbiR9IeSo9Hn9SizBxJsyrLBynNPF9bDqV5xSRpF0nbKwWEz1d63Pqb8rqZuexsSVMkTc6vv1dpaowNJJ0u6YbKts9Smh242p75knatLJ8h6dD8+0ck3STpWUpz1LxA0qaVdl6sNOHfVkqT6+2R122r9ATRDZRmWJ4r6fS6Om9QmpCw1u79lSbxW0fS25SyNU+vrLtL0ktyO7aVtHWT9m8haZnSl3yd3I5lkjarfP4LlZ5uPUnSenWfx2RJd9f2pbKvf1E6k/iRpJm9/l/jhx9+Rv9T6zck3SbpOZLWVZoja+v8fZ/ZZv87KS/vpTRlkJVmev+bpBfldUdLOq9S9z6Sbqost9VnK03Uuq9S5mIjpSdEX1gpO0dphoBtlLLct0r6Y97PSUrHim9Xyj8996vO27tbaQaDDfPyjrncCZIezH3qupJOljSvsp1WffdBSsfEf81tmKwWx4e8/RuVMjRTclteUdnWFXV/x9Ml/VjSJrnNF0k6Oa/bJdf92VzX5Lr33ilpl7rXlkp6aWX5OEn39vr/9fH29LoBHXzBDpS0uO61q5SmslitNLHeHLUZDDXY/umSTqv7Mj6zRXum5TIb13+xKmXma81gYoHSRH1S6ij2abLtqP2T5uUfSPpok7JvknR9XZ3vHeGzvKFWt6SfSTqiSbn69h8j6Zy6Mj+T9J78+xxJn2xR79lKc6C58tqrlC5xTlN6bPzNahHw8sMPP/39oyeCoY8pHdz3kPRzpQN2qMEJT5P+t2E/IOnCWp+lFCiskvTkvHyBpKObvG/EPrtS9oXVA3Xu246rLH9B0qWV5b21ZqB1iKRv5t/fXu2j6+o5QdLlleXnSlrd4rOt9t0HSVo4wt/i8eODUkZ+SaPPVWsfK60UeG1TeW0nSX/Ov+8i6WFJGzapt1EwdK7SCe9GSkHb7ZIe6vX/a+1nYC6TKWUgprtyXTQido6IaXldR/tie0fbv8zpwvuUxh5Nryu2qFJ+Xdun5JTlSqUvvBq8p1l920taGRG1bc5Q+mdopnrN9W9KE/3VJhb8vu27cjvObdXu/J53Ow1eXJFToM+rvGekdlRtLWn/2nbytl6hdBbUsO5KGz6f631r5G+GJEXE3Ejp1RWSjpD0DKWzSQCD7RxJ71A60M6urmiz/62V3dP2vHyJaYVSFmW6JEXEXyRdKWnffDloT0nfye9ru8+2/STbZ9pekMvOlTTN9rqVYvdUfl/dYHlqZbk6XqjTvn7D2nFuhL5bWruvb3V8mCFpQUQ80qItNZspZcmuq9T90/x6zZKIeLCNbdX8m9Ln9CelOeC+pxQ09YVBCoaulvSQUhq0mQeU/oA1T2tR9rtKKcAZEbGx0iWs+jvSqk+kfEeue1elNOnM/LoblG3k9ZIuqSwvUkq5durkXNfzI+LJShmzpu22vbWk/5T0QaXLcNOUsi+197RqR/0+LVLKDE2r/EyJiFNavEe2T1TqpF4XEStH2L9osD8ABkxELFAaSP16pYxAVTv9r2xvIOmHkk6VtHnuv35SV/ZspX5wf0lXR8Rd+fWR+uyqI5WGLOyY+9VXtSjbku31lC7n/Ty/NKq+vo2+W1q7v211fFgkaSs3Hmhdv52lSoHLdpW+fuOImNriPS1FxPJI43ufFhHbKcUfv+lkG900MMFQzhycKOnrtvezPdX2Onmg3JRc7AZJb8lR/rZKqcpmNpK0PCIetP1SpS9OKxspBWPLlAKuz9Stv0dSq2cY7aU17yz4hqRP2f5HJ8+3vekIbai1435JK2xvoTT2qJUpSv+0S6Q0GFHp7KLajqNsvzi3Y9v8JWy0T+dK2tv27vmsa8M8kG5LNWH735U+290iYlnduu1svzBva6pS6vkuSb8fYZ8ADIZDJL0mIh6oe73d/nd9pTEpSyQ9YntPSfWPJrlQaeb5I7RmBmqkPru+PauV+tVNJH1ipB1r4ZWSflc58btY0tNsfygPSt7I9o5tbGekvruRVseH3yiNXTrF9pTcf788r7tH0pa215ekiHhMKRA7zfZTc/1b2N69VeV5/zbMi+vnOqo3/2ya+/s9Jb1fUk+f81Q1MMGQJEXE5yR9WGnQ3F+V/oBnKo1luUppYNjD+fWzldOlTRwu6ZO2V0k6XmlcTiuzlcb83KU0eG5e3fpvSnpuTileWF3hdCfcc3Iba76Y67xM0sr8/nZuPT1R6Yt/n1Kmqf6Maw0RcatSkHG10ueyvVJaubb+fEmfVjpTW6XUsWySV58s6WN5n47Kl/j2kXSs0hd0kdKXrdX/0WeUBoH/yU88S6h2J8Xmks7L+3+H0pnbGyLi7yN/DAD6XUTcHhHXNljVVv8bEauULq/8QNK9SkHTj+vKrFbKHj1Da/aHI/XZVacr9b9Lc7mfjrRvLaxxS33eh92UxhUtVrpM9OqRNjJS391E0+NDRDya27Ct0o0udyoNypbSHdO3SFpse2l+7RilQePz8iW3y5WyZ63cphRUbqE0nnS10vAKKT1W5ial48zJkt4ZEbeMsL0JU8zcZE7PvHlU6U6phRNc91sl7RcRb53IegGgBLaPl/RPEXFgH7TlVqX+/tZetwXt69lDmnrgeUq3MDZ6GFS3rVDKWgEAxlG+rHWIpHf1QVvWlzSbQGjwDNRlstGyva+kX0o6JiIenuj6I+KyiLh6ousFgGFm+31Kl+svjYi5vW5PvjP2lJFLot8Uc5kMAACgkSIyQwAAAM0QDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKIRDAEAgKJ1NGv99OnTY+bMmV1qCtAfrrvuuqURsVmv2wF0E/05SjB//nwtXbrUI5XrKBiaOXOmrr322tG3ChgAthf0ug1At9GfowQ77LBDW+W4TAYAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIpGMAQAAIo2qdcNGBa2R/W+iBjnlgAAgE6QGQIAAEUjM9SB0WZ/OtkmmSIAACYWmSEAAFA0MkMtdCMT1E6dZIcAAJg4ZIYAAEDRyAw10IuMUKP6yRABANB9ZIbq9DoQqrLdV+0BAGAYEQwBAICicZksG88MTKvLW2R6AADoL2SGAABA0YrPDI0mUzOWgc317yVTBABAb5EZAgAARSs2MzTRGaGx4FZ7AAC6h8wQAAAoWrGZoU50MyNT2zZjhwAA6A2CIQAAxqidE1qGOvQvLpMBADBKncwUwBWA/kVmqAWieAAAhh/BEAAAHRptloe7g/sTl8kAAEDRyAwBANCm8Rr3Q4aovxQbDPXLPyAD6gAA6C0ukwEA0IZunLx2cjcauodgCAAAFK3Yy2S91smZQL9c0gMAjGw0Mwswhqi3CIYAAGih3aCmPpBhuqXBQTA0wcgIAQDQXxgzBABAF0UEJ7d9jmBoApEqBQCg/3CZrItGG/xwBgEAvdeLYQ22OQb0AJkhAABQNDJDXTCWy2GcEQDA4Oikz44Ihkv0KYIhAAAqCFjKQzA0Tsb65SEjBABAbzBmCAAAFI3M0BiREQIAYLARDAEA0CFOZIcLwdAokRECAGA4EAx1gAAIANBt3bqbjWNQcwygBgAARSMz1Aam1QCA4detjEy/PLeIqT6aIxgCABRhPIOSfglwOlVrN0HRmgiGWiAjBADA8CMYAgAMlUHN2kwkMkRrIhhqgIlWAQAoB8FQnU4DIYIfAOg9skEYC26tBwAARSMzNEpkhACg94Y5IzTa40wnnwljhxIyQwAAoGhkhgAAA2dQM0K1DEw32z8RdQwbgiEAwMDo5hyRnWx7EC4rdRIUlX65jGAoG7YvAQAAaA/BEABgIAzDo0+4dNWfCIY60I9fLAAAMDYEQwCAoTLaE9dhHXg8rPs1nri1HgAAFK34zBCRMgCgU6O5+6rd402rTA7DNbqj+GAIADAYImIgT2A7bXOr8qXfAt8tBENt4J8OAIDhRTAEAID6Y9hEu9mvTjJE/bBf/Y4B1AAAoGhkhgAAQ6VfxtVMREamfl9HW2evP6teIzMEAACKRmYIADCUJjJDNJ4TyI4my8O4oLEhMwQAAIpGZggAMNSqWZNmWaLRZlbGMyPUC72uv18QDAEABsZYBwoP0uWkbs4pRhC0JoKhIVb/BeKfHwCAtREMtaFfbtNsV7OzCNsDsw8A0Mqgz8TeSV88XvtK/98cA6gBAEDRyAx1YBAyRCOdOQzCPgBAuyYiQ9QvWahqv91OW+jn20dmCAAAFK34zNBoH27VbxF3r89YAKCXOs2adLrN2nInGZlGZcfr2NFvx6BBV3wwBAAYLt2cp2s0A5/R/wiGsn65JtyuQWknAAD9jmAIADCUyMygXQRDo9SLu7LIBgEAMP4Ihuq0O0CupptPeR7v4IezJAAA1sat9QAAoGhkhhoYy2DqdmZHbvf9Y0UmCACAkZEZAgAARSMz1MJYb7fvZJA1GSEAAHqDzBAAACgamaE2jFeGqBvIAgEAMDYEQx3oxtw3Y2kDAAAYOy6TAQCAopEZGqX6DA2XwgAAGExkhgAAQNHIDI2Tbsx6T0YIAIDuIzMEAACKRmZonHU60Wuj9wMAgIlDMNQFBDQAAAwOLpMBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiEQwBAICiOSLaL2wvkbSge80B+sLWEbFZrxsBdBP9OQrRVn/eUTAEAAAwbLhMBgAAikYwBAAAikYwBAAAikYwBAAAikYwBAAAikYwBAAAikYwBAAAikYwBAAAikYwBAAAivb/l0yQ5Iq5kuUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "\n",
    "def display_image_pair(first_item, second_item):\n",
    "    fig,(ax1, ax2) = plt.subplots(1, 2, figsize=(10, 10))\n",
    "    ax1.matshow(first_item[0], cmap='gray')\n",
    "    ax2.matshow(second_item[0], cmap='gray')\n",
    "    ax1.get_yaxis().set_visible(False)\n",
    "    ax1.get_xaxis().set_visible(False)\n",
    "    ax1.set_title(f'Train Sample:\\n{first_item[1]}')\n",
    "    ax2.get_yaxis().set_visible(False)\n",
    "    ax2.get_xaxis().set_visible(False)\n",
    "    ax2.set_title(f'Test Sample:\\n{second_item[1]}')\n",
    "    plt.subplots_adjust(wspace=1)\n",
    "    plt.show()\n",
    "    \n",
    "train_class = np.random.randint(train_set[1].shape[0])\n",
    "train_image = train_set[0][train_class][0].reshape(height, width)\n",
    "test_class = np.random.randint(len(test_set[1]))\n",
    "test_image = test_set[0][test_class][0].reshape(height, width)\n",
    "\n",
    "display_image_pair((train_image, train_set[1][train_class]), (test_image, test_set[1][test_class]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Omniglot Dataset\n",
    "\n",
    "The in memory data set is managed using OmniglotDataset class.\n",
    "\n",
    "The class load the data set using the OmniglotLoader class and split the train set in train and validation sets, leaving the test set unaltered."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading data from data/train.pickle\n",
      "Loading data from data/test.pickle\n"
     ]
    }
   ],
   "source": [
    "from omniglot_dataset import OmniglotDataset\n",
    "\n",
    "omniglot_dataset = OmniglotDataset()\n",
    "omniglot_dataset.load()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Each of the data sets tuple (train, validation and test) can be accessed using the corresponding class property/variable.\n",
    "Each data set type is stored in a tuple respectively with the first element being the data and the second the classes in the format 'alphabet/character' and they can be accessed using the corresponding instance property/variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Set:\n",
      "Shape = (673, 20, 105, 105, 1)\n",
      "Classes: ['Alphabet_of_the_Magi/character01' 'Alphabet_of_the_Magi/character02'\n",
      " 'Alphabet_of_the_Magi/character03' 'Alphabet_of_the_Magi/character04'\n",
      " 'Alphabet_of_the_Magi/character05']...['Korean/character36' 'Korean/character37' 'Korean/character38'\n",
      " 'Korean/character39' 'Korean/character40']\n",
      "\n",
      "Validation Set:\n",
      "Shape = (291, 20, 105, 105, 1)\n",
      "Classes: ['Latin/character01' 'Latin/character02' 'Latin/character03'\n",
      " 'Latin/character04' 'Latin/character05']...['Tifinagh/character51' 'Tifinagh/character52' 'Tifinagh/character53'\n",
      " 'Tifinagh/character54' 'Tifinagh/character55']\n",
      "\n",
      "Test Set:\n",
      "Shape = (659, 20, 105, 105, 1)\n",
      "Classes: ['Angelic/character01' 'Angelic/character02' 'Angelic/character03'\n",
      " 'Angelic/character04' 'Angelic/character05']...['ULOG/character22' 'ULOG/character23' 'ULOG/character24'\n",
      " 'ULOG/character25' 'ULOG/character26']\n"
     ]
    }
   ],
   "source": [
    "print('Train Set:')\n",
    "print(f'Shape = {omniglot_dataset.train_set[0].shape}')\n",
    "print(f'Classes: {omniglot_dataset.train_set[1][:5]}...{omniglot_dataset.train_set[1][-5:]}')\n",
    "print()\n",
    "print('Validation Set:')\n",
    "print(f'Shape = {omniglot_dataset.val_set[0].shape}')\n",
    "print(f'Classes: {omniglot_dataset.val_set[1][:5]}...{omniglot_dataset.val_set[1][-5:]}')\n",
    "print()\n",
    "print('Test Set:')\n",
    "print(f'Shape = {omniglot_dataset.test_set[0].shape}')\n",
    "print(f'Classes: {omniglot_dataset.test_set[1][:5]}...{omniglot_dataset.test_set[1][-5:]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class also has a property/variable with the data shape that uses the different data set types:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(105, 105, 1)"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "omniglot_dataset.data_shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The class also have two helper methods to retrieve random classes from a data set type:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Returned train classes = ['Arcadian/character09' 'Bengali/character19' 'Gujarati/character42'\n",
      " 'Burmese_(Myanmar)/character14']\n",
      "\n",
      "Returned validation classes = ['Latin/character04' 'Malay_(Jawi_-_Arabic)/character05'\n",
      " 'Mkhedruli_(Georgian)/character41' 'N_Ko/character29']\n",
      "\n",
      "Returned test classes = ['Syriac_(Serto)/character12' 'Malayalam/character17' 'Keble/character16'\n",
      " 'Kannada/character36']\n"
     ]
    }
   ],
   "source": [
    "train_classes = omniglot_dataset.get_data_classes(4)\n",
    "print(f'Returned train classes = {omniglot_dataset.train_set[1][train_classes]}')\n",
    "print()\n",
    "val_classes = omniglot_dataset.get_data_classes(4, data_type='val')\n",
    "print(f'Returned validation classes = {omniglot_dataset.val_set[1][val_classes]}')\n",
    "print()\n",
    "test_classes = omniglot_dataset.get_data_classes(4, data_type='test')\n",
    "print(f'Returned test classes = {omniglot_dataset.test_set[1][test_classes]}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And another method to get a random pair of images of the same/different class/es from a data set type, in this case the train data set type:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "first_class = train_classes[0]\n",
    "first_class_name = omniglot_dataset.train_set[1][first_class]\n",
    "first_image, second_image = omniglot_dataset.get_image_pair(first_class, data_type='train', same_class='true')\n",
    "display_image_pair((first_image.reshape(height, width), first_class_name), \n",
    "                   (second_image.reshape(height, width), first_class_name))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkMAAAD0CAYAAACYaSJ/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAGjRJREFUeJzt3XuUZFV96PHvD0YBQUBhIIoDk4ArIApEURCJoqKgiMbLQ1QEVAyYy72aKyhXUXkF0SiiBsWAhiAggo8EfKAiFw0iKEQQJPgiwwABFsMbhQj4u3/sXc6Zoqu6Z6a7q7v397NWremqc2qffWqq9v6d397nnMhMJEmSWrXKqCsgSZI0SgZDkiSpaQZDkiSpaQZDkiSpaQZDkiSpaQZDkiSpaQZDkyQiVo2IByJi41HXZWVExCURccCo6yFJWn4RcVlE7Dvqesw2zQZDNXDpPf4QEQ92nr9xecvLzEczc63MXLyC9fnriPhFRNwfEbdFxNcjYs0VKUuSWjTZ7Xqn3HEDjIh4e0T8sm7rtog4PyLWWNFtanrNG3UFRiUz1+r9HRGLgAMz88JB60fEvMx8ZCrqEhEvBY4Cds3MqyNiPWD3qdiWJM1Vy9uuT5aI2AU4gtKGX1Pb8NdM9XY1eZrNDI0nIo6NiC9FxBcj4n5g34h4fj1CuCcibo2IT0bE4+r68yIiI2JhfX5GXf6tmu35UUT86YDNPRf4YWZeDZCZd2bmaZn521rWqyPiqlrO4oh4f6eem9XtHhARN0fEXRHxtojYLiKuqXX9RGf9AyPiBxHx6Yi4NyL+IyJePORzODAiro+Iu+u+LFjJj1aSRqJOZ3h/RNwQEUsi4syIWLcuWzMizq5t6D0RcXlEPCkiPkZpo0+tWZ+PjVH0c4F/y8xr4I9t+Ocz88Fa9msj4uqIuC8iboyI93bqtHlEPBIRb42IWyLizoh4S+1vrq11OaGz/sERcVFEfLaWd11EvHDIPh9URx3uiohvRMRGk/Npzi0GQ8O9FjgLWAf4EvAI8A5gfeAFwK7AQUPe/wbg/cCTgcXAMQPWuwzYLSI+GBE7RMRqfcsfAPat9dgdeEdEvKpvnW2BTet6nwQOB14CPJMSyL2gs+4OwPV1P44BvtZrELoiYk/gMMoRznzg8vp59JZ/KyIOHbL/kjSTHAa8HNgReBrwMPDxuuxAymjJRpS28RDg95n5LuAnlCzTWvV5v8uAV0fEB2oQ8/i+5fdR+oN1Kf3KoRGxa2f5qsBWwJ8BbwY+BRwKvKi+/uaI2K6z/guBq4H1gOOBf4mItfsrFRH7AO+k9BsbAj8Fzugs/25EvHPAZ9UUg6HhLsnM8zPzD5n5YGb+JDMvz8xHMvMG4B8pX9ZBvpyZV2Tmw8CZwDZjrZSZFwN7Uo4uvgUsiYi/j4hV6vKLMvPaWo+rgbPH2O4xmfnfmflN4PfAGZl5R2beDFwC/EVn3VuBT2Xmw5l5FnAD8IoxqnYQcFxm/qIOER4LPK93ZJGZr8jMjw7Zf0maSQ4CDs/M/8rMhyjTE14XEUEJjOYDm9Y2/ie97Px46lDcPsB2wLcpbfiHO2349zLz57UN/3fgHB7bhh9d2/Dz6vPTa4ZpMXApy7bhN2Xmp2sbfjpwM7DLgP09NjN/Wfuho4AdI2LDWq+XZeaJE9nHuc5gaLibuk9qOvMbdXLcfcDRlCOIQW7r/P07YK1BK2bmNzLzVcCTgP8BvI1yhEA90rg4Iu6IiHspRzDr973/9s7TB4H+591t35zL3qH3RuCpY1RrE+Ckmqa9B1gC/IFyRCVJs0YNeBYA3+y0aT+l9IPrAZ8Dvg98uU45OC4iVp1o+Zl5XmbuRsn+7AW8HXhT3fYLIuL7nTb8AJZtwx/NzDs7z8dtw/s2P6wNP7mzv3dQRjhsw/sYDA2Xfc8/C1wLbJaZawMfAGJSN1iOHL4LXEwZ4oKSCfoKsCAz1wFOXcnt9v8QNgb+a4z1bgLempnrdh5rZOblK7FtSZp29QDwFuAlfW3a6pm5pGZlPpCZm1OGofaiZHvgsX3BsO38ITO/DfyApW34OZSpFr02/DSmrw0/YIw2/MqV2PacZDC0fJ4I3Av8NiK2YPh8oQmrk+v2rpP1IiK2B/6SMg7d2+5dmflQXbbPwMIm5ikRcUiUSd/7UOYaXTDGeicD76v7SkSsW+cRSdJsdDJwfO9EkIjYICJ2r3/vHBHPqENb91EyKI/W991Omc8zpojYMyL2qm1kRMQOlHmll9WM1FrAnbUN34ESaK2MBXUi9bwop/xvDHxnwP4eERF/Xuv5pIjYYyW3PScZDC2fdwH7A/dTskRfmqRy7wEOBn5N+RH+M2WuTq/8twMfinJW23spRxkr41JgS+Au4Ehgj8y8u3+lzDwXOAE4tw4L/ozOuHREfCci3r2SdZGk6fIR4ELgotqeXgo8uy7bCPhXSvt+LfBNlra1Hwf2i3JW7UfGKPdu4G+A31Da8M8DR2XmV2pG6mDgo3Wb7wbOXcn9+AFlDtFdwPuA12bmvf0rZeYXgX8Avlrb8KuAl/WW17PS/s9K1mVOiGWnjmiui4gDgX0zc6dR10WStHwi4mBgz8zcedR1mUvMDEmSpKYZDEmSpKY5TCZJkppmZkiSJDXNYEiSJDXNYGiAKDc+vaTz/IGIGHidiSmqQ0bEZtO5TUnSsiJiUUTsXP9+b0ScOs3bPzIizhh/Ta2oWRMM1dtR3D3GTUynRb1B3w2TVV69Rs/LJ6u8qdRtCFaijNUi4vNR7rJ8W/+1LSLiwIj4dQ06L4iIsS4tL2nEantwe0Ss2XntwIi4uPM8IuKwiPhVRDwYEYsj4vhe+x3lJs8P1MfDEfH7zvOTh2x7p3qQOLLrm2XmcZl54GSVF+Xm3JdOVnlTqT9JsBLlvDQiro+I30XE/4uITTrLNoqIf42Iu6LcFuXgld3eRMyKYCgiFlKuyJzAq4esN+H7yIxSbUSeQ7kPznRsb950bGfAtqNe0fVI4OmUe+W8GHh31Ls2R8SLgOOA1wBPBv4T+OJIKixpIuYB7xiy/JPAXwP7Ua6g/wrgJdSLGNabPK+VmWtRbmL9kd7zzBzW+e1PudDg/sMqN8o2bwW8knKBx2kx4v5gXkSsD3wVeD+lvb+CZS9gfAalD9gQ2A04LiJePOWVy8wZ/6DcA+yHlKshf73z+mnAZyhfpN8COwNrAB+j3LjuXsod29eo659LuXnqvZQreG7ZKWs94DzK1UN/DBxDuWt9b3lS7kkG5T/op3Xdm4AjO+strOvuDyym3Nz0fX3782rgvPr3qpSrSv+GcuXTKyn3r+lt82DgV5QrnJ7E0jMANwUuAu6s2zgTWLezjUXAeyhXjf5vSuN1eGc711GuWtqt19uA/+gsfzbwBcrNWR8EHgDeXdfdnnL11nuAq4GdOuVcDPxd/T97ENiMck+gl3fWOQY4u/79UeCkzrKn1n3fdNTfPR8+fCz7qG3L4ZSgZN362oHAxfXvp1NuY/G8vvctqG3RS/peP41yZ/XxtvuE2jbtA/we2LazrNfuvrW2uz+or+/Yaad69+ka2obX5W+i9CF3Uq7wvAjYuS47Ejijs+6wfuW02m5/o9b98v52Dfh34Nn17y2B79bP9nbgvZ1tngOcXsv5ed/+D2zbKTeF/SHlKtp3Accyfv+xgBKw3FHX+QdgC+Ch+n/7AHBPXXe12oYvrnU+maV97k6Um8q+p35GX6AEyZd2trUmpZ/YnHLbkgTmd5b/I/CFKf9ej/qHNcEf368plzp/DvAwsGHni3Yv5R4wqwCr1y/exZRLq68K7ACsVtd/C+UoZTXgROCqzjbOrl+2NSk317uFwcHQTsCz6ja3ql+Av+r7UZ5CCcy2pjQAW3TKOhk4qP59GHAN8OeUG/dtDazX2ebXKXdB3rh+MXetyzajXFZ9NWA+5Ud4Yl+DdVX9Uve+mHtRAo1VgNdRAsindJbdAjy31mMzYJNOWTt3yt6I8gN5ZS3rZfX5/Lr8YsoPY0tKEDa/7suGnTL2BK6pf38M+HRf+Qm8ZtTfPR8+fCz76LUHlM7y2PpaNxg6GLhxwHu/D3yo77XTmFgw9CbgVkq7fj7wyc6yXrt7OqUNX6O2mfcDrwceRzng3aauvxOD2/BnUDr7F9b29QTKfcoGBUPD+pXTKAHI82pbeCb1ILAuf0ptd6OWcSvltk+r1+fbdbb5UG1zVwU+BFzWKWdY235Arf//qnVYgyH9Ry3/akrwtGaty46dsi7p+385kZJIeHKt8/m9/+P6OT8CfLhuaw3gE8Bn+sq4Ftijvj+BDTrLTgF+OuXf61H/sCbwA9iREgCtX59fD/xt54t2emfdVSgR5tYTKHfd+qGvU//zHwY27yw/jgHB0BhlnQh8vO9H+bTO8h8D+3Se38jS7M8vGNDp13J27Dw/Bzh8wLp/1f3CUBqst4zzGVzV2zbwbeAdA9ZbxLLB0Hvoi9Tr+/evf18MHN1ZtqDuy+qd114GLKp/v5RydLJV/bF8lpKNev2ov38+fPhY9sHSYOiZlIPR+SwbDB1Bp6Pue+/ZwCl9r53GxIKhC1naYb+ecnD4uPq81+7+WWf9/wt8bYL71G3DP8CyAcualEzUmMFQXzl/7Fc6+3ZqZ/krges7z98KfK6zT2N2+nWbF3aePwN4cMj+dNv2A4DF4+z/H/sP4Pn1s503xnoHsGy/GJTAa9POa88H/rP+vVP97Lpt/+eA4/vK/SFLs3aXAJ+iBGHPpgSTv5jq7/VsmDO0P/CdzFxSn5/FsuPFN3X+Xp/yAf6mv5CIWLVO4PtNvWHdos575lMi5m5ZNw6qUERsVyd93RER91KOhNbvW+22zt+/o6T/iIhnAfdlZm9bC8aq7wTK2SAizo6IW+r+nDFGHbr7Q0TsFxFXRcQ9EXEPpTHrvWe8enRtAuzVK6eWtSPlKGesbT9Q/12789ralKM2MvN7wAeBr1A+90V12c0TrI+kaZaZ11Iy14f3LVrCsm1B11Pq8uUS5S7zL6ZkVqDcUHV1ynBXV7fdGdimjdOGP7VbTmb+lpL5HqucYf1Kz5hteNWdL7S8fcHqvfk/47Tt8Ni+YFj/sYCS2XtkSF165lOGL6/sbPuC+nrPHZn5UOf5AyzbF0CnPwDeCPxprfNnKP/nU94XzOhgKCLWAPYGXlTPQLoN+Ftg64jYuq6WnbcsoaQSNx2juDdQJujuTMkGLexthhIFP0L5EvRsPKRqZ1HSggsycx3KsFdMcLdeSRk/7rlpQH3H8yHKvm+VmWsD+45Rhz9+NnW2/inAIZRhuHUpqcnee4bVI/ue30TJDK3beayZmceP9Z7MvJuS/t26s3xryrh3b52TMvPpmbkBJSiaV+snaeb6IGWu4Uad1y4CFkTE87or1oBme+B7K7CdN1H6q/NrP3ADJRjar2+9bls1rE0b1obfSqcviIgnUIbYxjKsXxkqIh4HvIgyR2i8+g4rZ7y2HR7bhg/rP24CNh4w0bq/nCWU0ZgtO33BOlkmxg96z8/p9AX1hKJN6+tk5o2Z+arMnJ+Z21E++x8P/gQmx4wOhiipu0cpKcFt6mML4N947I+AzPwD8HnghIh4ao3an19P53wiZe7OnZRI9rjO+x6ljH8fGRFPiIhnMPxshScCd2XmQ/UH/4bl2KfdWPbMgVOBYyLi6fXMq60iYtAPr78ODwD3RMRGlLlHw6xJ+VLeARARb6YcPXTrcWhEPKfWY7PO6Y63A91rLJ0B7B4Ru9TPePV6yuvThmz/dOCIiHhSRGxOaUBPq3VZPSKeWbe7MWXC3CdqECVphsrMX1POBPrfndd+SQkuzoyI7WsbsSXlIOfCzLxwBTa1H3AUS/uBbShzTHYb0l6eCewcEXvXs5jWi4ht6rJhbfiXgVdFxI4R8XjgaAb3lQP7lQn4S+BnmXlfff514E8i4p1RLkXyxIjYbgLljNe2D6r3oP7jx5SA8PiIWLO2zy+oy24HnlY/l16fewrw8YjYoG5/o4jYZci2vwY8MyL2iIjVKcOSP8vM6+v7t6j7/viI2Bd4OWXe1pSa6cHQ/sA/ZebizLyt96DMbH8jJXvQ71DKhOSfUMYaP0zZz9MpQzC3UGbbX9b3vkMo6cvbKJ30Pw2p198AR0fE/ZT/yHMmsjMRsQ4lmOteU+KE+v7vUM5s+Bxl3sx4jqKMp95LyTR9ddjKmXkdZaLyjyhf6GdRxml7y8+lnAF2FiVd+S+UCXFQjiKOqGnQQ+sQ32soZ8HdQTmSOIzh36cPUlLAN1ImUf59Zl5Ql61et/sA5Yf4I8ppl5JmvqMpHXLXIZQDrDMov+sLKHMJ91jewiNie0rG5aRuP5CZ51FOrnn9WO/LzMWUTPy7KH3BVSzNSAxswzPz58D/pLRJt1LO5B00TDNevzLMMqfUZ+b9lLmUu1P6oV9RhgaHGq9tH2Bg/1GTA7tTJlkvpuz76+riiygZnNsiojfc+R7K/8NldcjtQsoJQYPqewfle/B3lM92O8oZgj27UDJ/d1OGL3et75lS3qh1GkXE3sCembn3qOsiSRqdiLiO0h9cN+q6aOZnhuaaeyinK0qSGlWHmU43EJo5zAxJkqSmmRmSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNm7c8K6+//vq5cOHCKaqKNDNceeWVSzJz/qjrIU0l23O1YNGiRSxZsiTGW2+5gqGFCxdyxRVXrHitpFkgIm4cdR2kqWZ7rhZsu+22E1rPYTJJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktS0eaOugCRJoxIR07q9zJzW7WlizAxJkqSmGQxJkqSmGQxJkqSmGQxJkqSmGQxJkqSmGQxJkqSmGQxJkqSmGQxJkqSmedFFSVKzJvMiiBO5gGNvHS++OLM0HwxN99VHB/GHIUnSaDhMJkmSmmYwNENExIzJUkmS1BKDIUmSJkFmOuVhljIYmmHMDkmSNL0MhiRJUtMMhiRJUtMMhmYgJ1NLkjR9DIYkSVLTmr/o4nTP/DfjI0nSzGJmSJKkaeZ0iJnFYEiSJDXNYEiSJDXNYGiaeYVSSZJmFoMhSZImkQe9s4/BkCRJaprBkCRJalrz1xmSJI3edJ9mPpuGsSbzs5lN+z2dzAxJkqSmmRmSJI2MFx4cbCo+m0Fltp4xMjMkSZKaZmZIkqQZZBTZsu42W8wSmRmSJElNMxiSJGmGmAlzqFq8iazDZJKkkekNybTW+fabifvfq1MLw2ZmhiRJUtPMDEmSRm4ysg8zMbsynhWt8/J8Xiv7ubSQITIzJEmSmmZmSJKkWWJFsjNjvWc2ZtGmkpkhSZLUNDNDkqRmTOe8l8k+U24y695f1kTqGBFzdt6QwZAkadaby8M+czUAmUkcJpMkSU0zMyRJ0gw0m4f0ZhszQ5IkqWlmhiRJTZgtc29mSz3nEjNDkiSpaWaGJEkasdmSDZqrt+YwMyRJkppmZkiSNGu1evaTJpeZIUmSBMy94a+JMhiSJElNc5hMkjSntZrt0MSZGZIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU0zGJIkSU2bN+oKjFpEjLtOZk5DTSRJ0iiYGZIkScDEEgRzkcGQJElqmsGQJElqmsGQJElqWvMTqCVJat1E5wrN1ROKzAxJkqSmGQxJkua0iGj2LClNjMNkkiQ1anmCxLk6RAYGQ3P6P1eSJI3PYTJJ0qyVmR7UaqUZDEmSpKY1P0wmSdJUmMmTtp0rtCwzQ5IkqWlmhiRJTYiIaclyzJWMUEsMhiRJzegFA5MVFM2m4GJ569rC8FiPw2SSJKlpZoYkSbNeZk55lmY2ZYG0fMwMSZKkppkZkiTNCb05LhPJ4AybOzTXMkCeRj8+M0OSJKlpZoYkSc2aa1kgrRgzQ5IkqWlmhmaZyTqKaXVcWNLctzxzh+aaFd3n1vsEgyFJkibBoIBiOoKyFgO/yWQwNAtMxZd8sq/CKknSbGUwNIMZ6UvSipvK4bKZciC5svs2U/Zj1JxALUmSmmZmSJI0p61shmgmZU88iWZqmBmSJElNayoz5BycpTwqkNQa2z0/g0GaCoYkSZptJvNAfmXKmsuBlMHQLDeXv5ySJE0H5wxJkjQFImJOTc+YS/vSz8zQLGMmSJKkyWUwNIMZ+EiSZpK5evcCh8kkSVLTzAxJkjQicy3DMluZGZIkSU1rKjM0kyLwuTwrX5Kk2aSpYEiSpNlmrHurTcfB/VgH7TMpqTCZHCaTJElNMzMkSdIsMN1ZmbmaBRqLmSFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktS0eaOugCRJM0FEzNptZuaklNMqM0OSJKlpBkOSJM1yETGSzNZcYTAkSZKaZjAkSZKaZjAkSZKaZjAkSZKa5qn1kiTNcp5av3LMDEmSpKaZGZrBeqdJGvFL0tSzrW2XmSFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0gyFJktQ0T60fkd4pnMPuMuxpnpIkTT0zQ5IkqWlmhkZsrAyRGSFJkqaPmSFJktQ0M0MzhNkgSZJGw8yQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqmsGQJElqWizPPbEi4g7gxqmrjjQjbJKZ80ddCWkq2Z6rERNqz5crGJIkSZprHCaTJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElNMxiSJElN+/8/MQoqL0wKQQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "first_class = train_classes[0]\n",
    "first_class_name = omniglot_dataset.train_set[1][first_class]\n",
    "first_image, second_image = omniglot_dataset.get_image_pair(first_class, data_type='train')\n",
    "display_image_pair((first_image.reshape(height, width), first_class_name), \n",
    "                   (second_image.reshape(height, width), f'NOT {first_class_name}'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Both methods will be used by the runner to create batchs of data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Siamese Neural Networks\n",
    "\n",
    "Siamese nets were first introduced in the early 1990s by Bromley and LeCun to solve signature verification as an image matching problem [Bromley et al., 1993](https://papers.nips.cc/paper/769-signature-verification-using-a-siamese-time-delay-neural-network.pdf). \n",
    "\n",
    "A siamese neural network consists of twin networks which accept distinct inputs but are joined by an energy function at the top. This function computes some metric between the highestlevel feature representation on each side. \n",
    "\n",
    "The parameters between the twin networks are tied. Weight tying guarantees that two extremely similar images could not possibly be mapped by their respective networks to very different locations in feature space because each network computes the same function. \n",
    "\n",
    "Also, the network is symmetric, so that whenever we present two distinct images to the twin networks, the top conjoining layer will compute the same metric as if we were to present the same two images but to the opposite twins.\n",
    "\n",
    "![alt text](./images/siamese_networks.png \"Siamese Neural Networks\")\n",
    "\n",
    "The paper uses the weighted L1 distance between the twin feature vectors h1 and h2 combined with a sigmoid activation, which maps onto the interval [0, 1]. Thus a cross-entropy objective is a natural choice for training the network.\n",
    "\n",
    "The CNN architecture of each network has 9 layers, including Max Pooling and Convolutional layers of different filter sizes.\n",
    "These 9 layers work as feature extractor/selector for the siamese architecture.\n",
    "\n",
    "![alt text](./images/feature_extractor.png \"Feature Extractor Network\")\n",
    "\n",
    "All layers in the feature extractor are initialized with weights having 0 mean 0.01 standard deviation, and bias with a mean value of 0.5 and a standard deviation of 0.01, except for the first convolutional layer (the input layer), which its bias are not initialized. L2 regularization of 0.0004 is also used for all layers except for the last one that uses 0.002.\n",
    "\n",
    "Adam optimizer with the default configuration has been used to compile the model."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Siamese Net\n",
    "\n",
    "The SiameseNet class create an instance of the siamese networks with the specific data input that is passed in the constructor, in this case the one defined by omniglot_dataset.data_shape property.\n",
    "\n",
    "The instance created has a model property/variable whichs contains the keras model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "left_encoder_input (InputLayer) (None, 105, 105, 1)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "right_encoder_input (InputLayer (None, 105, 105, 1)  0                                            \n",
      "__________________________________________________________________________________________________\n",
      "encoder (Sequential)            (None, 4096)         38947648    left_encoder_input[0][0]         \n",
      "                                                                 right_encoder_input[0][0]        \n",
      "__________________________________________________________________________________________________\n",
      "l1_distance (Lambda)            (None, 4096)         0           encoder[1][0]                    \n",
      "                                                                 encoder[2][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_2 (Dense)                 (None, 1)            4097        l1_distance[0][0]                \n",
      "==================================================================================================\n",
      "Total params: 38,951,745\n",
      "Trainable params: 38,951,745\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "from siamese_net import SiameseNet\n",
    "\n",
    "siamese_net = SiameseNet(omniglot_dataset.data_shape)\n",
    "siamese_net.model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Following is the layers of one of the feature extractor networks:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<keras.layers.convolutional.Conv2D at 0xb19f7a080>,\n",
       " <keras.layers.pooling.MaxPooling2D at 0xb1f194a20>,\n",
       " <keras.layers.convolutional.Conv2D at 0xb1f1bef60>,\n",
       " <keras.layers.pooling.MaxPooling2D at 0xb1f1dbac8>,\n",
       " <keras.layers.convolutional.Conv2D at 0xb1f1f52b0>,\n",
       " <keras.layers.pooling.MaxPooling2D at 0xb1f207da0>,\n",
       " <keras.layers.convolutional.Conv2D at 0xb1f2239e8>,\n",
       " <keras.layers.core.Flatten at 0xb1f24fdd8>,\n",
       " <keras.layers.core.Dense at 0xb1f26e278>]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "siamese_net.model.layers[2].layers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model loss, metrics and optimizer configuration:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model loss function: binary_crossentropy\n",
      "Model metric functions: ['accuracy']\n",
      "Model optimizer: <keras.optimizers.Adam object at 0xb1f3815c0>\n"
     ]
    }
   ],
   "source": [
    "print(f'Model loss function: {siamese_net.model.loss}')\n",
    "print(f'Model metric functions: {siamese_net.model.metrics}')\n",
    "print(f'Model optimizer: {siamese_net.model.optimizer}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One Shot Runner\n",
    "\n",
    "The OneShotRunner class run one-shot experiments on a model using a specific data set.\n",
    "\n",
    "The model and data set are passed as parameters when creating the runner instance, in our case we create the runner object with a omniglot data set and a siamese network model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from one_shot_runner import OneShotRunner\n",
    "\n",
    "one_shot_runner = OneShotRunner(omniglot_dataset, siamese_net.model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A runner instance has two methods, train and predict in order to run experiments.\n",
    "\n",
    "Train method, as the name suggests, train the model using the data set for a series of iterations, using a specific n-way value and evaluating the training performance on the validation set with a specific number of validations.\n",
    "\n",
    "All those parameters are configurable in the method signature, the default configuration is for experiments on one-shot 20-way tasks with epochs of 10000 iterations each and 50 validation on the validation data set type for each evaluation."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_shot_runner.train(number_way=20, number_iterations=10000, number_validations=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The runner reports/save the following parameters during the training process:\n",
    "\n",
    "- Report training loss and accuracy every 100 iterations\n",
    "- Report validation loss and accuracy every 1000 iterations\n",
    "- Save training/evaluation state (loss and accuracy) every 1000 iterations\n",
    "- Save the model weights at the end of the process\n",
    "\n",
    "All the parameters are stored at the ./model folder within the training_data.pickle and model_weights.h5 files respectively.\n",
    "\n",
    "Below is the output that the training process reports while executing:\n",
    "\n",
    "*Start training for 10000 iterations with 50 validations per each one-shot 20-way task  \n",
    "iteration 100, loss = 3.24, accuracy = 78.12  \n",
    "iteration 200, loss = 2.52, accuracy = 90.62  \n",
    "iteration 300, loss = 2.20, accuracy = 84.38  \n",
    "iteration 400, loss = 1.92, accuracy = 90.62  \n",
    "iteration 500, loss = 1.72, accuracy = 87.50  \n",
    "iteration 600, loss = 1.48, accuracy = 96.88  \n",
    "iteration 700, loss = 1.44, accuracy = 90.62  \n",
    "iteration 800, loss = 1.45, accuracy = 84.38  \n",
    "iteration 900, loss = 1.23, accuracy = 96.88  \n",
    "iteration 1000, loss = 1.13, accuracy = 93.75  \n",
    "evaluation at iteration 1000, loss = 0.33, accuracy = 60.00  \n",
    "Saving training data at iteration 1000  \n",
    "iteration 1100, loss = 1.08, accuracy = 90.62  \n",
    "...  \n",
    "...  \n",
    "...  \n",
    "iteration 9900, loss = 0.11, accuracy = 100.00  \n",
    "iteration 10000, loss = 0.12, accuracy = 100.00  \n",
    "evaluation at iteration 10000, loss = 0.06, accuracy = 94.00  \n",
    "Saving training data at iteration 10000  \n",
    "Saving model weights*\n",
    "\n",
    "Saving the model state/data allow us to start the experiment in an specific training point, to do so, we just need to enable the corresponding flag parameter in the train method, which is disabled by default."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_shot_runner.train(preload_state=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Preloading model weights  \n",
    "Preloading training data  \n",
    "Start training for 10000 iterations with 50 validations per each one-shot 20-way task  \n",
    "...  \n",
    "...  \n",
    "...  *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Every iteration, the model is feeded with batches of 32 samples from the train set, each sample is 2 omniglot images.\n",
    "\n",
    "Half of the batch samples uses a pair of images that belong to the same class and the other half belong to different classes.\n",
    "The classes of the first image of each sample are different and picked randomly from the train data set.\n",
    "The batch is shuffle before feeding the model.\n",
    "\n",
    "A 8 samples batch would be as displayed below.\n",
    "\n",
    "**NOTE THAT THE METHOD IS PUBLIC JUST TO SHOW HOW A TRAINING BATCH IS, IT SHOULD BE ENCAPSULATED (PRIVATE) IN THE RUNNER CLASS AND THE BATCH_SIZE BE CONFIGURABLE**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "inputs, labels = one_shot_runner.get_train_batch(batch_size=8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Labels: [1. 1. 0. 1. 0. 0. 0. 1.]\n"
     ]
    }
   ],
   "source": [
    "from functools import reduce\n",
    "\n",
    "def display_batch(first_input, second_input):\n",
    "    first_input = first_input.squeeze()\n",
    "    second_input = second_input.squeeze()\n",
    "    first_plot = reduce((lambda x, y: np.append(x, y, axis=1)), first_input)\n",
    "    second_plot = reduce((lambda x, y: np.append(x, y, axis=1)), second_input)\n",
    "    final_plot = np.append(first_plot, second_plot, axis=0)\n",
    "    fig,(ax1) = plt.subplots(1, 1, figsize=(15, 15))\n",
    "    ax1.matshow(final_plot, cmap='gray')\n",
    "    ax1.get_yaxis().set_visible(False)\n",
    "    ax1.get_xaxis().set_visible(False)\n",
    "    plt.show()\n",
    "display_batch(*inputs)\n",
    "print(f'Labels: {labels}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Every evaluation is made using a one-shot 20-way task per each validation. An example of a one-shot 20-way task input is as follow:\n",
    "\n",
    "**NOTE THAT AGAIN, THE METHOD IS PUBLIC JUST TO SHOW HOW A ONE-SHOT TASK BATCH IS.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x1080 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Labels: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "inputs, labels = one_shot_runner.get_one_shot_batch(batch_size=20, data_type='val')\n",
    "display_batch(*inputs)\n",
    "print(f'Labels: {labels}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Predict method, as the name suggests, predicts the output of a one-shot 20-way task using the trained model. \n",
    "\n",
    "A number of predictions are done for all data set types (train, validation and test) with a number of validation per predictions.\n",
    "\n",
    "Each validation, the model is feeded with a one-shot N-way task batch generated from the corresponding data set type (train, validation or test) as showed above.\n",
    "\n",
    "All those parameters are configurable in the method signature, the default configuration is for predictions on one-shot 20-way tasks with 100 predictions and 50 validations for each prediction."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_shot_runner.predict(number_way=20, number_iterations=1000, number_validations=50)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The runner reports/save the following parameters during the prediction process:\n",
    "\n",
    "- Report prediction iteration every 10 iterations\n",
    "- Save prediction accuracy (train, validation and test) at the end of the process\n",
    "\n",
    "The predictions are stored at the ./model folder within the prediction_data.pickle.\n",
    "\n",
    "Below is the output that the prediction process reports while executing:\n",
    "\n",
    "*Start predictions for 100 iterations with 50 validations per each one-shot 20-way task  \n",
    "Predictions at iteration 10 finished  \n",
    "Predictions at iteration 20 finished  \n",
    "Predictions at iteration 30 finished  \n",
    "Predictions at iteration 40 finished  \n",
    "Predictions at iteration 50 finished  \n",
    "Predictions at iteration 60 finished  \n",
    "Predictions at iteration 70 finished  \n",
    "Predictions at iteration 80 finished  \n",
    "Predictions at iteration 90 finished  \n",
    "Predictions at iteration 100 finished  \n",
    "Saving prediction data to model/predictions_data.pickle  *\n",
    "\n",
    "The prediction can be done using a specific model state, we just need to enable the corresponding flag parameter in the predict method, which is disabled by default."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "one_shot_runner.predict(preload_state=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "*Preloading model weights  \n",
    "Start predictions for 100 iterations with 50 validations per each one-shot 20-way task  \n",
    "...  \n",
    "...  \n",
    "...*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Experiments\n",
    "\n",
    "With all the described components in place we can start running one-shot N-way task experiments.\n",
    "\n",
    "### Default Configuration Experiment\n",
    "\n",
    "The most basic experiment would be to simply create all the components with the default configuration and train the network for an epoch and test the quality of the predictions.\n",
    "\n",
    "The code would be as follow:\n",
    "\n",
    "```python\n",
    "#Create omniglot data set instace\n",
    "omniglot_dataset = OmniglotDataset() \n",
    "\n",
    "#Create siamese network with an specific data input shape and default configuration\n",
    "siamese_net = SiameseNet(omniglot_dataset.data_shape)\n",
    "\n",
    "#Create a one-shot runner instance with a specific data set and model\n",
    "one_shot_runner = OneShotRunner(omniglot_dataset, siamese_net.model)\n",
    "\n",
    "#Run the experiment, this time with the default configuration\n",
    "one_shot_runner.train()\n",
    "one_shot_runner.predict()\n",
    "```\n",
    "\n",
    "Once the experiment finishes, all the data generated can be processed and displayed using the Utils and RunnerVisualizer components.\n",
    "\n",
    "The Utils class has two static methods to save/read data in pickle format and is used by some components to store/load data.\n",
    "\n",
    "The RunnerVisualizer class has three static methods to display the loss, accuracy and predictions data respectively.\n",
    "\n",
    "For the experiment with the default configuration we can see that the network doesn't converge and the results are very bad for any of the data set types, having more or less an accuracy average of 8% for all of them."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from utils import Utils\n",
    "from runner_visualizer import RunnerVisualizer\n",
    "\n",
    "def report_results(model_folder):\n",
    "    training_data = Utils.read_data(f'./{model_folder}/training_data.pickle')\n",
    "    training_loss, training_accuracy, validation_loss, validation_accuracy = training_data\n",
    "    predictions_data = Utils.read_data(f'./{model_folder}/predictions_data.pickle')\n",
    "    train_accuracy, val_accuracy, test_accuracy = predictions_data\n",
    "\n",
    "    RunnerVisualizer.display_loss(training_loss);\n",
    "    print('\\n\\n\\n')\n",
    "    RunnerVisualizer.display_accuracy(validation_accuracy);\n",
    "    print('\\n\\n\\n')\n",
    "    data = [train_accuracy, val_accuracy, test_accuracy]\n",
    "    RunnerVisualizer.display_predictions(data)\n",
    "    \n",
    "model_folder = 'experiments/model_default_parameters'\n",
    "report_results(model_folder)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Learning Rate = 0.00005 Experiment\n",
    "\n",
    "The code would be as follow:\n",
    "\n",
    "```python\n",
    "#Create omniglot data set instace\n",
    "omniglot_dataset = OmniglotDataset() \n",
    "\n",
    "#Create siamese network with an specific data input shape\n",
    "siamese_net = SiameseNet(omniglot_dataset.data_shape)\n",
    "\n",
    "#Configure the model with the desired parameters, learning rate of 0.00005 this time\n",
    "from keras import backend as K\n",
    "K.set_value(siamese_net.model.optimizer.lr, 0.00005)\n",
    "\n",
    "#Create a one-shot runner instance with a specific data set and model\n",
    "one_shot_runner = OneShotRunner(omniglot_dataset, siamese_net.model)\n",
    "\n",
    "#Run the experiment, this time during 40000 iterations (4 epochs)\n",
    "one_shot_runner.train(number_iterations=40000)\n",
    "one_shot_runner.predict()\n",
    "```\n",
    "\n",
    "For the experiment with the above configuration we can see that the network this time converges and the results are much better, achieving an average accuracy of 82% for the test set, but with a worst case scenario of 68%.\n",
    "\n",
    "We can also see that the cost is decreasing each epoch but the validation accuracy did not improve during the third epoch but it did a lot at the end of the last epoch.\n",
    "\n",
    "The model has overfitted the train set which achieves an accuracy average of 96%, which suggets that there is still plenty room for improvement and also knowing that author of the paper obtained a test result of 92%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "\n",
      "\n",
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "model_folder = 'experiments/model_lr'\n",
    "report_results(model_folder)"
   ]
  }
 ],
 "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}