code_session_version/lab-11-0-cnn-basics-session.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import absolute_import\n",
    "from __future__ import division\n",
    "from __future__ import print_function"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 3, 3, 1)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x17b2997cb38>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQ8AAAD8CAYAAABpXiE9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAADi5JREFUeJzt3X/MnWV9x/H3ZxQqUWarhdGUIpI1ds4tEZ8g6mKaqQk2hi6RJfiHgtE0Osl00WSoCSYmy9Q/XGYwkqpEWAySidG61BgEHC4LjEoKpTSVlmThSRtAsEWiU8q+++O52c4O5+nz9Dr3c84pvl/Jybl/XOe+vlxNPr3uXzRVhSSdrN+bdgGSTk2Gh6QmhoekJoaHpCaGh6QmhoekJmOFR5JXJLktycPd99pF2j2XZE/32TlOn5JmQ8Z5ziPJF4CnqupzSa4B1lbV345o90xVvWyMOiXNmHHD4wCwpaqOJFkP/LiqXjOineEhvciMGx5Hq2rNwPovquoFpy5JjgN7gOPA56rqu4scbzuwHeClL33pGzZv3txc24vdc889N+0SZt6zzz477RJm3r59+35eVWe3/HbVUg2S/Ag4d8SuT59EP+dX1eEkFwJ3JNlbVYeGG1XVDmAHwNzcXO3evfskuvjdcvTo0WmXMPMee+yxaZcw8zZv3vyfrb9dMjyq6u2L7UvyWJL1A6ctjy9yjMPd9yNJfgy8HnhBeEg6dYx7q3YncGW3fCXwveEGSdYmWd0trwPeAjw0Zr+Spmzc8Pgc8I4kDwPv6NZJMpfka12bPwJ2J7kfuJOFax6Gh3SKW/K05USq6kngbSO27wY+2C3/O/An4/Qjafb4hKmkJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmhoekJoaHpCaGh6QmvYRHkkuTHEhyMMk1I/avTnJLt/+eJBf00a+k6Rk7PJKcBnwZeCfwWuA9SV471OwDwC+q6g+BfwA+P26/kqarj5nHxcDBqnqkqn4LfAvYNtRmG3Bjt/xt4G1J0kPfkqakj/DYADw6sD7fbRvZpqqOA8eAV/bQt6Qp6SM8Rs0gqqENSbYn2Z1k9xNPPNFDaZJWSh/hMQ9sHFg/Dzi8WJskq4CXA08NH6iqdlTVXFXNnX322T2UJmml9BEe9wKbkrw6yRnAFcDOoTY7gSu75cuBO6rqBTMPSaeOVeMeoKqOJ7ka+CFwGnBDVe1L8llgd1XtBL4O/FOSgyzMOK4Yt19J0zV2eABU1S5g19C2aweW/wv4yz76kjQbfMJUUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUhPDQ1ITw0NSE8NDUpNewiPJpUkOJDmY5JoR+69K8kSSPd3ng330K2l6Vo17gCSnAV8G3gHMA/cm2VlVDw01vaWqrh63P0mzoY+Zx8XAwap6pKp+C3wL2NbDcSXNsLFnHsAG4NGB9XngjSPavTvJW4GfAX9TVY8ON0iyHdgOcM4553D77bf3UN6L04EDB6Zdwsw7dOjQtEt4Uetj5pER22po/fvABVX1p8CPgBtHHaiqdlTVXFXNrVmzpofSJK2UPsJjHtg4sH4ecHiwQVU9WVW/6Va/Cryhh34lTVEf4XEvsCnJq5OcAVwB7BxskGT9wOplwP4e+pU0RWNf86iq40muBn4InAbcUFX7knwW2F1VO4G/TnIZcBx4Crhq3H4lTVcfF0ypql3ArqFt1w4sfxL4ZB99SZoNPmEqqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpSS/hkeSGJI8neXCR/UnypSQHkzyQ5KI++pU0PX3NPL4BXHqC/e8ENnWf7cBXeupX0pT0Eh5VdRfw1AmabANuqgV3A2uSrO+jb0nTMalrHhuARwfW57tt/0+S7Ul2J9l99OjRCZUmqcWkwiMjttULNlTtqKq5qppbs2bNBMqS1GpS4TEPbBxYPw84PKG+Ja2ASYXHTuB93V2XS4BjVXVkQn1LWgGr+jhIkpuBLcC6JPPAZ4DTAarqemAXsBU4CPwKeH8f/Uqanl7Co6res8T+Aj7SR1+SZoNPmEpqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGpieEhqYnhIamJ4SGrSS3gkuSHJ40keXGT/liTHkuzpPtf20a+k6enlH7oGvgFcB9x0gjY/qap39dSfpCnrZeZRVXcBT/VxLEmnhr5mHsvxpiT3A4eBT1TVvuEGSbYD2wHOPPNMrrvuugmWd2rZu3fvtEuYeYcOHZp2CS9qkwqP+4BXVdUzSbYC3wU2DTeqqh3ADoC1a9fWhGqT1GAid1uq6umqeqZb3gWcnmTdJPqWtDImEh5Jzk2Sbvnirt8nJ9G3pJXRy2lLkpuBLcC6JPPAZ4DTAarqeuBy4MNJjgO/Bq6oKk9LpFNYL+FRVe9ZYv91LNzKlfQi4ROmkpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmowdHkk2Jrkzyf4k+5J8dESbJPlSkoNJHkhy0bj9SpquPv6h6+PAx6vqviRnAT9NcltVPTTQ5p3Apu7zRuAr3bekU9TYM4+qOlJV93XLvwT2AxuGmm0DbqoFdwNrkqwft29J09PrNY8kFwCvB+4Z2rUBeHRgfZ4XBoykU0gfpy0AJHkZcCvwsap6enj3iJ/UiGNsB7YDnHnmmX2VJmkF9DLzSHI6C8Hxzar6zogm88DGgfXzgMPDjapqR1XNVdXc6tWr+yhN0grp425LgK8D+6vqi4s02wm8r7vrcglwrKqOjNu3pOnp47TlLcB7gb1J9nTbPgWcD1BV1wO7gK3AQeBXwPt76FfSFI0dHlX1b4y+pjHYpoCPjNuXpNnhE6aSmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmhgekpoYHpKaGB6SmowdHkk2Jrkzyf4k+5J8dESbLUmOJdnTfa4dt19J07Wqh2McBz5eVfclOQv4aZLbquqhoXY/qap39dCfpBkw9syjqo5U1X3d8i+B/cCGcY8rabalqvo7WHIBcBfwuqp6emD7FuBWYB44DHyiqvaN+P12YHu3+jrgwd6K68c64OfTLmKA9ZzYrNUDs1fTa6rqrJYf9hYeSV4G/Cvwd1X1naF9vw/8d1U9k2Qr8I9VtWmJ4+2uqrleiuvJrNVkPSc2a/XA7NU0Tj293G1JcjoLM4tvDgcHQFU9XVXPdMu7gNOTrOujb0nT0cfdlgBfB/ZX1RcXaXNu144kF3f9Pjlu35Kmp4+7LW8B3gvsTbKn2/Yp4HyAqroeuBz4cJLjwK+BK2rp86UdPdTWt1mryXpObNbqgdmrqbmeXi+YSvrd4ROmkpoYHpKazEx4JHlFktuSPNx9r12k3XMDj7nvXIE6Lk1yIMnBJNeM2L86yS3d/nu6Z1tW1DJquirJEwPj8sEVrOWGJI8nGfkMThZ8qav1gSQXrVQtJ1HTxF6PWObrGhMdoxV7haSqZuIDfAG4plu+Bvj8Iu2eWcEaTgMOARcCZwD3A68davNXwPXd8hXALSs8Lsup6Srgugn9Ob0VuAh4cJH9W4EfAAEuAe6ZgZq2AP8yofFZD1zULZ8F/GzEn9dEx2iZNZ30GM3MzAPYBtzYLd8I/MUUargYOFhVj1TVb4FvdXUNGqzz28Dbnr8NPcWaJqaq7gKeOkGTbcBNteBuYE2S9VOuaWJqea9rTHSMllnTSZul8PiDqjoCC/+xwDmLtHtJkt1J7k7Sd8BsAB4dWJ/nhYP8v22q6jhwDHhlz3WcbE0A7+6mwN9OsnEF61nKcuudtDcluT/JD5L88SQ67E5pXw/cM7RramN0gprgJMeoj+c8li3Jj4BzR+z69Ekc5vyqOpzkQuCOJHur6lA/FTJqBjF8L3s5bfq0nP6+D9xcVb9J8iEWZkZ/voI1ncikx2c57gNeVf/3esR3gRO+HjGu7nWNW4GP1cB7Xs/vHvGTFR+jJWo66TGa6Myjqt5eVa8b8fke8NjzU7fu+/FFjnG4+34E+DELKdqXeWDwb+3zWHiRb2SbJKuAl7OyU+Yla6qqJ6vqN93qV4E3rGA9S1nOGE5UTfj1iKVe12AKY7QSr5DM0mnLTuDKbvlK4HvDDZKsTbK6W17HwtOtw//fkHHcC2xK8uokZ7BwQXT4js5gnZcDd1R3xWmFLFnT0PnyZSyc007LTuB93R2FS4Bjz5+OTsskX4/o+jnh6xpMeIyWU1PTGE3iCvQyrwi/ErgdeLj7fkW3fQ74Wrf8ZmAvC3cc9gIfWIE6trJwNfoQ8Olu22eBy7rllwD/DBwE/gO4cAJjs1RNfw/s68blTmDzCtZyM3AEeJaFv0E/AHwI+FC3P8CXu1r3AnMTGJ+larp6YHzuBt68grX8GQunIA8Ae7rP1mmO0TJrOukx8vF0SU1m6bRF0inE8JDUxPCQ1MTwkNTE8JDUxPCQ1MTwkNTkfwBRARJelRPLdQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sess = tf.InteractiveSession()\n",
    "image = np.array([[[[1],[2],[3]],\n",
    "                   [[4],[5],[6]], \n",
    "                   [[7],[8],[9]]]], dtype=np.float32)\n",
    "print(image.shape)\n",
    "plt.imshow(image.reshape(3,3), cmap='Greys')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1 filter (2,2,1,1) with padding: VALID\n",
    "\n",
    "weight.shape = 1 filter (2 , 2 , 1, 1)\n",
    "![image](https://cloud.githubusercontent.com/assets/901975/24833375/c0d9c262-1cf9-11e7-9efc-5dd6fe0fedb0.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image.shape (1, 3, 3, 1)\n",
      "weight.shape (2, 2, 1, 1)\n",
      "conv2d_img.shape (1, 2, 2, 1)\n",
      "[[12. 16.]\n",
      " [24. 28.]]\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAM0AAAC7CAYAAADGxxq1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAACXdJREFUeJzt3V+opPV9x/H3p1r1wnazuk1cTFIj1bQmLcQs1iYQpUYwUtxALJibaFEW20qhVzUIKeSmmpvSYNqwSUO1F0bqRbMphhJrlgTKWpei2cRgXKXBZZeYmLJlaZt0028v5kk6nMzZc77Oc2bmrO8XDPPMPL/z/L6MfHz+7A++qSokbd7PLbsAabsxNFKToZGaDI3UZGikJkMjNc0VmiQXJflykheG953rjPtxkmeG14F55pSWLfP8O02STwA/qKr7k9wL7KyqP5kx7lRVXThHndLKmDc0zwPXV9WJJLuBg1X19hnjDI3OGvPe07ypqk4ADO9vXGfcBUkOJzmU5INzzikt1bkbDUjyBHDJjF33NeZ5a1UdT3I58GSSI1X14oy59gH7ho/vbhz/de/CCz2Rd506der7VfVL3b/bMDRV9f719iX5bpLdU5dnr6xzjOPD+0tJDgLvAn4mNFW1H9g/HNtFcQ179uxZdgnbzsGDB7/zWv5u3suzA8Dtw/btwBfWDkiyM8n5w/Yu4L3Ac3POKy3NvKG5H7gxyQvAjcNnkuxJ8tlhzK8Bh5M8C3wFuL+qDI22rQ0vz86kql4Fbpjx/WHgrmH7n4Ffn2ceaZW4IkBqMjRSk6GRmgyN1GRopCZDIzUZGqnJ0EhNhkZqMjRSk6GRmgyN1GRopCZDIzUZGqnJ0EhNhkZqMjRSk6GRmgyN1GRopCZDIzUZGqnJ0EhNhkZqMjRSk6GRmgyN1GRopKZRQpPkpiTPJzk6NKxdu//8JI8O+59KctkY80rLMHdokpwDfAr4AHAV8OEkV60Zdifw71X1K8CfAw/MO6+0LGOcaa4BjlbVS1X1I+DzwN41Y/YCDw3bjwE3JMkIc0sLN0ZoLgVenvp8bPhu5piqOg2cBC4eYW5p4ebqhDaYdcZY22R2M2PWdneWVtIYZ5pjwFumPr8ZOL7emCTnAjuAH6w9UFXtr6o9VWWrYq2sMULzNHBFkrclOQ+4jUnX52nTXaBvBZ6sKluea1ua+/Ksqk4nuQf4R+Ac4HNV9c0kHwcOV9UB4K+Bv01ylMkZ5rZ555WWZYx7GqrqceDxNd99bGr7v4HfHWMuadlcESA1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UtOiujvfkeR7SZ4ZXneNMa+0DHO32pjq7nwjk45nTyc5UFXPrRn6aFXdM+980rItqruzdNYYo6nTrO7Ovzlj3IeSvA/4NvDHVfXyjDE/deWVV7J///4Rynt9uO6665ZdwraTzOqfvLExzjSb6dz8ReCyqvoN4AngoZkHSvYlOZzk8MmTJ0coTRrfQro7V9WrVfXD4eNngHfPOtB0d+cdO3aMUJo0voV0d06ye+rjLcC3RphXWopFdXf+oyS3AKeZdHe+Y955pWVZVHfnjwIfHWMuadlcESA1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UtNY3Z0/l+SVJN9YZ3+SfHLo/vz1JFePMa+0DGOdaf4GuOkM+z8AXDG89gF/NdK80sKNEpqq+iqTZk3r2Qs8XBOHgDes6Y4mbRuLuqeZ1QH60gXNLY1qUaHZTAdouztrW1hUaDbsAA12d9b2sKjQHAA+MjxFuxY4WVUnFjS3NKpRGtUmeQS4HtiV5Bjwp8DPA1TVp5k0sb0ZOAr8J/B7Y8wrLcNY3Z0/vMH+Av5wjLmkZXNFgNRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYZGajI0UpOhkZoMjdRkaKQmQyM1GRqpydBITYvq7nx9kpNJnhleHxtjXmkZRmm1waS784PAw2cY87Wq+p2R5pOWZlHdnaWzxiLvaX4rybNJvpTkHQucVxpVJk3KRjhQchnwD1X1zhn7fhH436o6leRm4C+q6ooZ4/YB+4aP7wRm3iMt2S7g+8suYh2rWtuq1vX2qvqF7h8tJDQzxv4bsKeq1v0hkxyuqj2jFDeiVa0LVre2s62uhVyeJbkkSYbta4Z5X13E3NLYFtXd+Vbg95OcBv4LuK3GOsVJC7ao7s4PMnkk3bH/tVe0pVa1Lljd2s6quka7p5FeL1xGIzWtTGiSXJTky0leGN53rjPux1PLcQ5sYT03JXk+ydEk987Yf36SR4f9Tw1PD7fcJuq6I8n3pn6juxZU10ZLqZLkk0PdX09y9YrU1V/iVVUr8QI+Adw7bN8LPLDOuFMLqOUc4EXgcuA84FngqjVj/gD49LB9G/DoitR1B/DgEv77vQ+4GvjGOvtvBr4EBLgWeGpF6rqeyT+VbPqYK3OmAfYCDw3bDwEfXGIt1wBHq+qlqvoR8Hkm9U2brvcx4IafPFZfcl1LURsvpdoLPFwTh4A3JNm9AnW1rVJo3lRVJwCG9zeuM+6CJIeTHEqyVcG6FHh56vOx4buZY6rqNHASuHiL6unUBfCh4RLosSRv2eKaNmuztS9Da4nXWKucNyXJE8AlM3bd1zjMW6vqeJLLgSeTHKmqF8ep8KdmnTHWPmbczJixbWbOLwKPVNUPk9zN5Gz421tc12Ys4/fajH8Ffrn+f4nX3wM/s8Rr2kJDU1XvX29fku8m2V1VJ4bT9ivrHOP48P5SkoPAu5hc54/pGDD9f+g3A8fXGXMsybnADrZ+pfeGdVXV9EqLzwAPbHFNm7WZ33Thquo/prYfT/KXSXbVGZZ4rdLl2QHg9mH7duALawck2Znk/GF7F/Be4LktqOVp4Iokb0tyHpMb/bVP6qbrvRV4soY7yy20YV1r7hNuAb61xTVt1gHgI8NTtGuBkz+5HF+m17TEa9FPWc7wlONi4J+AF4b3i4bv9wCfHbbfAxxh8tToCHDnFtZzM/BtJmex+4bvPg7cMmxfAPwdcBT4F+DyBf1OG9X1Z8A3h9/oK8CvLqiuR4ATwP8wOavcCdwN3D3sD/Cpoe4jTBbsrkJd90z9XoeA92x0TFcESE2rdHkmbQuGRmoyNFKToZGaDI3UZGikJkMjNRkaqen/AJ8IgtACvr2/AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print(\"imag:\\n\", image)\n",
    "print(\"image.shape\", image.shape)\n",
    "weight = tf.constant([[[[1.]],[[1.]]],\n",
    "                      [[[1.]],[[1.]]]])\n",
    "print(\"weight.shape\", weight.shape)\n",
    "conv2d = tf.nn.conv2d(image, weight, strides=[1, 1, 1, 1], padding='VALID')\n",
    "conv2d_img = conv2d.eval()\n",
    "print(\"conv2d_img.shape\", conv2d_img.shape)\n",
    "conv2d_img = np.swapaxes(conv2d_img, 0, 3)\n",
    "for i, one_img in enumerate(conv2d_img):\n",
    "    print(one_img.reshape(2,2))\n",
    "    plt.subplot(1,2,i+1), plt.imshow(one_img.reshape(2,2), cmap='gray')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1 filter (2,2,1,1) with padding:SAME\n",
    "![image](https://cloud.githubusercontent.com/assets/901975/24833381/fd01869e-1cf9-11e7-9d59-df08c7c6e5c4.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image.shape (1, 3, 3, 1)\n",
      "weight.shape (2, 2, 1, 1)\n",
      "conv2d_img.shape (1, 3, 3, 1)\n",
      "[[12. 16.  9.]\n",
      " [24. 28. 15.]\n",
      " [15. 17.  9.]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print(\"imag:\\n\", image)\n",
    "print(\"image.shape\", image.shape)\n",
    "\n",
    "weight = tf.constant([[[[1.]],[[1.]]],\n",
    "                      [[[1.]],[[1.]]]])\n",
    "print(\"weight.shape\", weight.shape)\n",
    "conv2d = tf.nn.conv2d(image, weight, strides=[1, 1, 1, 1], padding='SAME')\n",
    "conv2d_img = conv2d.eval()\n",
    "print(\"conv2d_img.shape\", conv2d_img.shape)\n",
    "conv2d_img = np.swapaxes(conv2d_img, 0, 3)\n",
    "for i, one_img in enumerate(conv2d_img):\n",
    "    print(one_img.reshape(3,3))\n",
    "    plt.subplot(1,2,i+1), plt.imshow(one_img.reshape(3,3), cmap='gray')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3 filters (2,2,1,3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "image.shape (1, 3, 3, 1)\n",
      "weight.shape (2, 2, 1, 3)\n",
      "conv2d_img.shape (1, 3, 3, 3)\n",
      "[[12. 16.  9.]\n",
      " [24. 28. 15.]\n",
      " [15. 17.  9.]]\n",
      "[[120. 160.  90.]\n",
      " [240. 280. 150.]\n",
      " [150. 170.  90.]]\n",
      "[[-12. -16.  -9.]\n",
      " [-24. -28. -15.]\n",
      " [-15. -17.  -9.]]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print(\"imag:\\n\", image)\n",
    "print(\"image.shape\", image.shape)\n",
    "\n",
    "weight = tf.constant([[[[1.,10.,-1.]],[[1.,10.,-1.]]],\n",
    "                      [[[1.,10.,-1.]],[[1.,10.,-1.]]]])\n",
    "print(\"weight.shape\", weight.shape)\n",
    "conv2d = tf.nn.conv2d(image, weight, strides=[1, 1, 1, 1], padding='SAME')\n",
    "conv2d_img = conv2d.eval()\n",
    "print(\"conv2d_img.shape\", conv2d_img.shape)\n",
    "conv2d_img = np.swapaxes(conv2d_img, 0, 3)\n",
    "for i, one_img in enumerate(conv2d_img):\n",
    "    print(one_img.reshape(3,3))\n",
    "    plt.subplot(1,3,i+1), plt.imshow(one_img.reshape(3,3), cmap='gray')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## MAX POOLING\n",
    "![image](https://cloud.githubusercontent.com/assets/901975/23337676/bd154da2-fc30-11e6-888c-d86bc2206066.png)\n",
    "\n",
    "![image](https://cloud.githubusercontent.com/assets/901975/23340355/a4bd3c08-fc6f-11e6-8a99-1e3bbbe86733.png)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 1, 1, 1)\n",
      "[[[[4.]]]]\n"
     ]
    }
   ],
   "source": [
    "image = np.array([[[[4],[3]],\n",
    "                    [[2],[1]]]], dtype=np.float32)\n",
    "pool = tf.nn.max_pool(image, ksize=[1, 2, 2, 1],\n",
    "                    strides=[1, 1, 1, 1], padding='VALID')\n",
    "print(pool.shape)\n",
    "print(pool.eval())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## SAME: Zero paddings\n",
    "\n",
    "![image](https://cloud.githubusercontent.com/assets/901975/23340337/71b27652-fc6f-11e6-96ef-760998755f77.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 2, 2, 1)\n",
      "[[[[4.]\n",
      "   [3.]]\n",
      "\n",
      "  [[2.]\n",
      "   [1.]]]]\n"
     ]
    }
   ],
   "source": [
    "image = np.array([[[[4],[3]],\n",
    "                    [[2],[1]]]], dtype=np.float32)\n",
    "pool = tf.nn.max_pool(image, ksize=[1, 2, 2, 1],\n",
    "                    strides=[1, 1, 1, 1], padding='SAME')\n",
    "print(pool.shape)\n",
    "print(pool.eval())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "mnist = keras.datasets.mnist\n",
    "class_names = ['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']\n",
    "#mnist = keras.datasets.fashion_mnist\n",
    "#class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat', 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "(train_images, train_labels), (test_images, test_labels) = mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_images = train_images.astype(np.float32) / 255.\n",
    "test_images = test_images.astype(np.float32) / 255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x17b3f1662b0>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "img = train_images[0]\n",
    "plt.imshow(img, cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"Conv2D_3:0\", shape=(1, 14, 14, 5), dtype=float32)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#sess = tf.InteractiveSession()\n",
    "\n",
    "img = img.reshape(-1,28,28,1)\n",
    "W1 = tf.Variable(tf.random_normal([3, 3, 1, 5], stddev=0.01))\n",
    "conv2d = tf.nn.conv2d(img, W1, strides=[1, 2, 2, 1], padding='SAME')\n",
    "print(conv2d)\n",
    "sess.run(tf.global_variables_initializer())\n",
    "conv2d_img = conv2d.eval()\n",
    "conv2d_img = np.swapaxes(conv2d_img, 0, 3)\n",
    "for i, one_img in enumerate(conv2d_img):\n",
    "    plt.subplot(1,5,i+1), plt.imshow(one_img.reshape(14,14), cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Tensor(\"MaxPool_2:0\", shape=(1, 7, 7, 5), dtype=float32)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pool = tf.nn.max_pool(conv2d, ksize=[1, 2, 2, 1], strides=[\n",
    "                        1, 2, 2, 1], padding='SAME')\n",
    "print(pool)\n",
    "sess.run(tf.global_variables_initializer())\n",
    "pool_img = pool.eval()\n",
    "pool_img = np.swapaxes(pool_img, 0, 3)\n",
    "for i, one_img in enumerate(pool_img):\n",
    "    plt.subplot(1,5,i+1), plt.imshow(one_img.reshape(7, 7), cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}