{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "imperial-battlefield",
   "metadata": {},
   "source": [
    "# Tensorflow"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "genetic-candle",
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import cv2 as cv\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "liquid-butterfly",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([2, 2, 9])"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "subject = tf.constant([\n",
    "    [[1, 2, 3, 4, 5, 6, 7, 8, 9], [1, 2, 3, 4, 5, 6, 7, 8, 9]], \n",
    "    [[1, 2, 3, 4, 5, 6, 7, 8, 9], [1, 2, 3, 4, 5, 6, 7, 8, 9]]\n",
    "])\n",
    "subject.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "diagnostic-connection",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(213, 320, 3)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(1, 213, 320, 3)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "im = cv.imread('sheep.png')\n",
    "print(im.shape)\n",
    "im = im / 255\n",
    "im = np.expand_dims(im, 0)\n",
    "im.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "extraordinary-beads",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([1, 213, 320, 6])"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer = tf.keras.layers.Conv2D(6, 1)\n",
    "result = layer(im)\n",
    "result.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "changing-eligibility",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorShape([1, 213, 320, 10])"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "seq = tf.keras.Sequential(\n",
    "    layers=[tf.keras.layers.Conv2D(6, 1),\n",
    "           tf.keras.layers.Conv2D(10, 1)]\n",
    ")\n",
    "\n",
    "seq.compile()\n",
    "res = seq(im)\n",
    "res.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "statewide-steal",
   "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": "3.8.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}