Facial_Expression_Recognition_Model.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "a36d01fa",
   "metadata": {},
   "source": [
    "### Importing all the necessary Python packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "ac899aa1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np \n",
    "import pandas as pd \n",
    "import cv2\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5004dca0",
   "metadata": {},
   "source": [
    "### Importing all the necessary Deep Learning packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1f100770",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tensorflow.keras.utils import to_categorical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "3a897be8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Dense , Activation , Dropout ,Flatten\n",
    "from keras.layers.convolutional import Conv2D\n",
    "from keras.layers.convolutional import MaxPooling2D\n",
    "from keras.metrics import categorical_accuracy\n",
    "from keras.models import model_from_json\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "from keras.optimizers import *\n",
    "from keras.layers import BatchNormalization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "b320bca0",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('fer2013.csv') #Reading the dataset using Pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "59d45d77",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>emotion</th>\n",
       "      <th>pixels</th>\n",
       "      <th>Usage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>70 80 82 72 58 58 60 63 54 58 60 48 89 115 121...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>151 150 147 155 148 133 111 140 170 174 182 15...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>231 212 156 164 174 138 161 173 182 200 106 38...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>24 32 36 30 32 23 19 20 30 41 21 22 32 34 21 1...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>6</td>\n",
       "      <td>4 0 0 0 0 0 0 0 0 0 0 0 3 15 23 28 48 50 58 84...</td>\n",
       "      <td>Training</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   emotion                                             pixels     Usage\n",
       "0        0  70 80 82 72 58 58 60 63 54 58 60 48 89 115 121...  Training\n",
       "1        0  151 150 147 155 148 133 111 140 170 174 182 15...  Training\n",
       "2        2  231 212 156 164 174 138 161 173 182 200 106 38...  Training\n",
       "3        4  24 32 36 30 32 23 19 20 30 41 21 22 32 34 21 1...  Training\n",
       "4        6  4 0 0 0 0 0 0 0 0 0 0 0 3 15 23 28 48 50 58 84...  Training"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fa716330",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2304"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(df.iloc[0]['pixels'].split()) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "04ae0a9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Number of pixels is 48*48"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "59cbcb6a",
   "metadata": {},
   "outputs": [],
   "source": [
    "label_map = ['Anger', 'Neutral', 'Fear', 'Happy', 'Sad', 'Surprise']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "96e953aa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2304"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Converting the pixels from strings to integer values\n",
    "img = df.iloc[0]['pixels'].split()\n",
    "img = [int(i) for i in img]\n",
    "type(img[0])\n",
    "len(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "37f3ffcf",
   "metadata": {},
   "outputs": [],
   "source": [
    "img = np.array(img)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c3dde346",
   "metadata": {},
   "outputs": [],
   "source": [
    "img = img.reshape(48,48)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "738ee477",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(48, 48)"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "d1297904",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, '0')"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAEICAYAAACZA4KlAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAAiSUlEQVR4nO2db6yeVZnur5vSFrBCKaWltLX/qPJHBCIoaAUCGtAZwQ9qRscTJiHywZnEycxxxHOSkzPJOVG/jExyjnNCjmaqTgbnX4SQGUfkdJhMNEWgLQINpQiFlpZSaAsVxXazzof9btLnWtfe773ftu/e23X9EkLX6nqeZz3ree4++772fd8rSikwxvz2c9JUT8AYMxxs7MY0go3dmEawsRvTCDZ2YxrBxm5MIxyTsUfEjRHxZERsj4jbj9ekjDHHnxj09+wRMQvANgAfAbATwM8AfKaU8sR4x8yZM6ecdtppnb6TTur+exMR1XEjIyMTticx5077zTff7DtGwdd/29veVo2ZM2dO1cf3qtae+9QYvr66D9XXbz7cBoBZs2ZVfSeffHKnPXv27ON27kHgNXr99derMYcPH+601XNWa8ZrfeTIkb5j1DPLPI/M+8l9fK3Dhw9jZGREvsQnq84k7wOwvZTyi95E7wJwM4Bxjf20007DunXrOn1vf/vbO231Arz22mud9iuvvNJ3cuo8/MKpl4JfXHWeAwcOdNpXXHFFNWbVqlVV3ymnnNJpq3+0+KX8zW9+0/f6v/rVr6oxqq/ffObNm1eNOf3006u+hQsXdtqLFi2qxvBznTt3bjVm/vz5nbZ6ufkfFmVsvGZbtmypxuzZs2fC8wLAr3/966rv4MGDnfZLL71Ujdm/f/+E8wHq56HeK373Dh06VI3hPl6zHTt2VMeMcSw/xi8F8PxR7Z29PmPMNOSEC3QRcVtEPBQRD6mvlDFmOByLse8CsPyo9rJeX4dSyp2llMtLKZcrP9YYMxyOxWf/GYC1EbEKo0b+ewA+O9EBb775ZuWDsi+lfDIWIZSve+aZZ3ba7I8Ctf+r/DbWB5T/9dGPfrTTVj47C5GAvjcm8w/iG2+8MenzKoGMfUS1Zmo+7NurMdyXESwV7NuqZ8/ay2WXXVaN2bBhQ6e9a1f1XZLPjOeors/rpt4ZRgmE7Ner95PXnu99IoF5YGMvpRyJiD8C8K8AZgH4dinl8UHPZ4w5sRzLlx2llH8G8M/HaS7GmBOII+iMaYRj+rJPlpGREbz88sudPvY5zjjjjOo4/p2t8pvYv1G/M2Vfin9fDQALFizotD/84Q9XYy699NJOW/mjgwaMsB+tfDA+twpq4eMy58n49apP+Za8Jmo9eE4qGIXHKH2Cr8/vCwCsXbu2037yySerMepe2Y/m37sD9Tus1oPvTcVBsF6kNATWB/haE/ns/rIb0wg2dmMawcZuTCPY2I1phKEKdEeOHKlEMRYclCjBwf8qQYDFjYwgdP7551djbrjhhk77nHPOqcawkJUV6HicSvzg4zIJEwq+VyW+sbiTWTNFJqNNiVZ8/0p45fOohBo+txL6Lrjggk77Rz/6UTVGCbYcxKKENV4jdR8sLKp3OBNO3m/NJspi9ZfdmEawsRvTCDZ2YxphqD47UPsc7LuoghIcIKMKKrDfqMZcfPHFnfbVV19djVm8ePGE51UoH0358XzvGV83E5yTKdSROU755xl9QF2f10T57DxHNSYDX18F3px99tmd9rnnnluNefjhh6u+pUu7JRpUwA4Xr1AJRRn4OC7uAdQ+OWtVE+EvuzGNYGM3phFs7MY0go3dmEYYetYbBy5wkIQSls4666xOm8UWoBbk1qxZU4255JJLOm1VAppR4tOg5bf53jKlgtV6sPilhDVe16nemlsFjGTKTWeExgynnnpqp7169epqzI9//OOqjwWwFStWVGNeffXVTjuTYaiCg1iI3rdvXzWGs0IzJarH8JfdmEawsRvTCDZ2Yxph6EE1DPsumd1FVLDBO9/5zk575cqV1Rj22wbdfol9MhUMovy2THJK5jxMxh/PXCsT+KLmpOY4mQoqYyj/M/M8BkHt2KOqG+3evbvTVjoP60Uq0CUTQMQVkrmqE1An4kxmffxlN6YRbOzGNIKN3ZhGsLEb0whDF+g4uIDLO2dKBaugGg68UUELma2EWKTJZIJlxCegvjclppyo/fAy+7xnBMvscZmtpQbJclNBTryuSujjMaoCkZojb9Gszp3ZjjkzR14zJUTzuVkMdClpY4yN3ZhWsLEb0whD9dlLKZWPzn6K8mXYt1NbRLGvq4JB2LfMJJmoQAseo6q5ZIJRlD7BfcqvZb8tE4yi5shjMgk1QH3/me24VFVWRq1ZJlGKx6gtkzkRR/nnqgrNCy+80GkrP5rfkYxelAlyUnPkdczoQG9ds+8VjTG/FdjYjWkEG7sxjWBjN6YRhh5UkxG3GM4qUplH3KcCG1jIUgLZIKWjMyKaGqcEMSUuMZlSxf222QLqe80Ex6i+N954oxrD96a29cqIS5ktohj1PDLrqp79L3/5y05bCbaZ58FrlpmPgq/vSjXGmAobuzGN0NfYI+LbEbE3Ih47qm9BRNwXEU/1/n/mROcwxkw9GZ/9rwH8LwDfOarvdgD3l1K+FhG399pf7neiiKj8Gw5AWLBgQXUc93FFDyCXVJGp6pEJNMlUiVU+Wca/4jGq6glvkaWCQfg+eIsioF575XurNeItstT1OYglmyzEZLbDYj9eaTH9zgvoCriZbZQz12OfXWkhmUStflWDjykRppTy7wBeoe6bAazv/Xk9gE/0O48xZmoZ1GdfXEoZK861B8DiiQYbY6aeY/7VWymlRMS4AbkRcRuA24DjV/DfGDN5BrW+FyNiCQD0/r93vIGllDtLKZeXUi4f1G8zxhw7g37Z7wFwC4Cv9f5/d+agk046qRLouMLM2rVrq+OWLVvWaasAjcw+5ix4ZI5RAg0LUlyierxz871nftJRAhlnPilhjc+dEYTUGBYDgTqwQ93roUOHOm1VgYfXUa01X0uJYZkts/g4JaCqe2XhV60RP391HkZ9+DIZj/2yRI9JoIuIvwXwUwDvioidEXErRo38IxHxFIAP99rGmGlM3y97KeUz4/zV9cd5LsaYE4gVM2MaYaiJMCeffHLlo/M2TWorXQ6qGbRSKftEmUQY5Y+yX6SCH1Qf+1fK189sEaUq9TDsk6rqKer6jNqC6ODBg5220gz4OLUe/BxVcA4fp3ztQfSaV17h0JH6vtQclT/Oz2jQLbt4jkrn6FdtydVljTE2dmNawcZuTCPY2I1phKEKdLNnz6623VmxYkWnrbZ2YqEis4+4yjDLlIlmAUSJJC+++GKnzQEkgC6dzOKSElMWLlzYaasAIhbouJIPUM9bnYePy5bE5ky8V199tRrDfWqOu3bt6rRZvAWAefPmddqZACb1XPttmwToNeJAH3VuXjf1fvJ7lam4kxHxJoO/7MY0go3dmEawsRvTCDZ2YxphyiPouMSUivTKlJtmUUQJdBzppaKxWGxTQhtHh6nIKyXsZSLv9u3b13dMZk9uXg/OHARqMVRFsKm9zfh6SkhiQUxFnnEUW6bklHpmLOKp+fCc1fNRAh2/Myozb5DozcwcM0xmj3t/2Y1pBBu7MY1gYzemEYbqs5900knVNk0cJKGCONiXy/g2mUwoFSDB/rfyNTPZa8rXZn9T+Z+8bRNvPwTUfn1mS6KdO3f2HaOCWlTZ7uXLl/cdw3qAulcOvFHBSazhKJ+Z/W/24YGc7qO2FeMS3CpTMlNhZqJ90yczpl9WpLPejDE2dmNawcZuTCPY2I1phKEKdBFRCSUsKGT21lIBESyKKKEvk53EopUSv1gge+aZZ6oxSijJ7L/GKIEwsx84i19KWOLsOXWvSiDkDDYVjMMiJguPALB79+5OW5W3YsH0+eefr8bwuqr5cFCRevaqbxAxOBPkpGBRV52Hn31mn7kx/GU3phFs7MY0go3dmEYYqs9eSqmCKzIJApntfdhvUmPYl8ucRwXHcMCI8ptUgAb7+irQhe9VJQZdd911nbbSDDKlnLlPJb2o+2DNRAUV8XHPPvtsNeaHP/xhp63KVrNmoZ4Z+8OXXXZZ3/moNVPvHp9bBQfxOqo5ZoJf+Nlngmwy9vPWHPqezRjzW4GN3ZhGsLEb0wg2dmMaYcoFOhZgVAZVRtxgIUUJWyxeZPa/5tLOQC1I7dixoxqjgoP67a0NAM8991ynfdVVV1Vj1q1b12mrICMWoJTQxs9i6dKl1RjeZw8AlixZ0mkvWrSoGsPr9oMf/KAas3nz5k5bVQViMfTaa6+txjB79+6t+ljIUnu9ZSruKIGOz53JsMtkZar5TCaIhvGX3ZhGsLEb0wg2dmMaYag++8jISJWgwckPyo/N7JmeCfRgH10FmnDgjfIj2Zc699xzqzGPPfZY1cdz+tSnPlWN4eup6jE871WrVlVjOPFD7enOQT6Z4Bg1J+Xrsx+vdAWueKOSZZg1a9ZUfZdeemmn/cADD1Rj2NfOVMUBan0o42sP6o/zGmUCbyZTxclfdmMawcZuTCPY2I1phL7GHhHLI2JDRDwREY9HxBd7/Qsi4r6IeKr3//oX5MaYaUNGoDsC4E9LKY9ExNsBPBwR9wH4AwD3l1K+FhG3A7gdwJcnOtHIyEi1dVFme59MNhALEypghgWYQQQQoK7ooiqjXHTRRVUfZ35xpRagLoOsMsE4aEOJbyysKfGJA5jUuqrjONAms/f7jTfeWI3hwBsVDMPP8ZprrqnGMDfccEPVt23btk5bVeVRYmQmC3IQ1LvHzzXznk9my6i+X/ZSyu5SyiO9P78GYCuApQBuBrC+N2w9gE+kr2qMGTqT8tkjYiWAywBsBLC4lDL2adoDYPE4x9wWEQ9FxEPqq22MGQ5pY4+IeQD+EcAfl1I6vywvoz9vyKz5UsqdpZTLSymXZ2KGjTEnhlRQTUTMxqih/00p5Z963S9GxJJSyu6IWAKgdriIw4cP44UXXuj0cfXSzLa9KvCG/Rs1hs+t/vHhMcpn5YAM5cephB4OvmH9AqgTHTJBRqoyCvufSsPgAB5V3VXdPwffZLbZvvDCC/teX20rzRqGqsjL81FBRps2beq0M9oQkPOJB6kcq+A5qfeT5zOZ6jYZNT4AfAvA1lLKXxz1V/cAuKX351sA3N3vXMaYqSPzZf8ggP8E4OcRsbnX918AfA3A30XErQB2APj0CZmhMea40NfYSyn/AWC8n2WuP77TMcacKBxBZ0wjDDXrbc6cOZUIw+KKEhhYOFLZWTwmI74pYYXFLhUwwmNU9RB1fRZTVDAMizQZYUdtB8XXUnPkcytxVJ2bq/CoOfIacZANAKxevbrT3rVrVzWGsyQVHJyjxNHMdlgq0GaQDLaMqJcRmTPX9vZPxpgKG7sxjWBjN6YRhuqzz507t6o0wv638pvYb8xUGVG+DAdxKJ+I/abM9tDqWipog8+d2Uoos7VwxkdUc8wkeWTmqCrpZu6DE4jOOeecaszKlSs7bXWvGS2G71+dJ+P/Kk1pkOQYdQzPSY3JbHM2Hv6yG9MINnZjGsHGbkwj2NiNaYShb//EIggLcqp0c0YA4eOUSJM5Dwc7qEwwDrRRWV8qaILFrky1kgyZgBklfGayCZVAx/er1lWJdgyLn6riD2dFKsGURV61hrxGSpxV65ipDMPHqTXLZMbx+qtjeF0ze7iP4S+7MY1gYzemEWzsxjSCjd2YRhiqQBcRlbiTiSpjlCCWiY7LZNhlBA8WjVRmnBKJWFxS98FkRCMFRxmqElj9ShyNdy0WjtQz4+ehMhUZtWa8tipijNdVCYY8ZyV+qftngTZTfly9Q5OJdJvoPIPsPffWHCY9A2PMjMTGbkwj2NiNaYSh++z9gkaUv8U+kfKb2G9Tvi77X8pvY/9T+Vp8bnUttZUQ30dmuyEFX+/111+vxhw8eHDCNpDbxkr18bqpajZ8HyoYpt95x+tj+BkpDSGT9aaupYKqGOXr9yOT0ZYZw/bk/dmNMTZ2Y1rBxm5MI9jYjWmEoQp0ChYUMmWZVUYVB3Eo0YivlRFflGjDYpO61qD7f2X2rOuXOQgABw4c6LQ5EAio91nPBLUAdYCMEt+4T43hdVP3wXNSz4zvTb0fHGSUFQM58EmVRMsE9QxSJlqNGaTU+Bj+shvTCDZ2YxrBxm5MIww9qIYDIDIBCeyXKD+FfXbe1xuofStVFYf9YRX4wr5/ZhspNU75luynqfXh41SSC/vs7J+rPnUtdW98/5nAI/XMeG2VPzxI2e6XXnqpGrNv374Jzzse7LMPsh2UIlPxRl2L5+1KNcaYChu7MY1gYzemEWzsxjTC0INq+mVDZcoyDxpowgEag+5rzuJKRugDcoEmmfNwEMnLL79cjWHRSgXMsNCn5pPJIFNrzaKdEt+UiNrv+pk90p555plqDGcGqntV7xWvW2YdM2uWEQgzFYkypa7H8JfdmEawsRvTCH2NPSJOiYgHI2JLRDweEX/e618VERsjYntEfD8i+v9MaoyZMjI++xsAriulHIqI2QD+IyL+BcCfAPhGKeWuiPg/AG4F8FcTnUgF1WSCbDj4IuMjZ/Z5V34TJ35kfMTJBDZM9jgVeMM+u/IROdBm165d1Rhex/3791dj1Dpy9ZoLLrigGrNo0aJOWyXisD6zZMmSagz79SoRhtdx69at1Rh+PzJ6CZDTizI6T2Y7LGbQ92o8+n7Zyyhj6srs3n8FwHUA/qHXvx7AJ47rzIwxx5WUzx4RsyJiM4C9AO4D8DSAA6WUsX/CdgJYekJmaIw5LqSMvZQyUkq5FMAyAO8DcH72AhFxW0Q8FBEPqcKIxpjhMCk1vpRyAMAGAFcBmB8RY07vMgC1Uzh6zJ2llMtLKZerQgjGmOHQV6CLiLMBHC6lHIiIUwF8BMDXMWr0nwRwF4BbANyduSAH1WS2IGLBQwU2sEilBBEWmzJ7uGcyuhQquIHPldneJ7Ovuspo27ZtW6f9k5/8pBrz7LPPdtp79uypxqg5Llu2rNPesmVLNYbvVQli11xzzYTnBXJBNZzl9tRTT1Vj+J1R71nmeWSyGdWzn0xFmYmOOZbtnzJq/BIA6yNiFkZ/Evi7Usq9EfEEgLsi4n8A2ATgW+mrGmOGTl9jL6U8CuAy0f8LjPrvxpgZgCPojGmEoSbClFIqH5QDXZS/k9mSKUNm2ya+lkrMYd9O+XrK3+I+dVxm+18OalH+55o1azptVb2Ffc3zzjuvGqOCYbgKjgpyWrlyZaf9oQ99qBrDPnrmeah15SAanh9Q36vSQpT/y76+GsPrn9kyepAto4DB333AX3ZjmsHGbkwj2NiNaQQbuzGNMHSBLlNimOmXKQdMLrhgjEx2UqacrxIVM6JdJrBDCYS8/ZJaw6VLu6kK1157bTWGRTsltKky1VyWWQX1rFu3rtNWmXGciaeeB6+/EgwfffTRqo/JBEJlyATMqDHcN+j+7BbojDF9sbEb0wg2dmMaYag++5tvvln5XOyDqO2WMpVSGeXrZqqFsD+ufKRMpZrM9j6Z6yu/nqu1KN8uo42w73/66adXYxYvXlz1cRUadRzPW1XT4etntut+8sknqzHbt2/vtDNVg7PPJxNUM8h2zAp+RtktqrL4y25MI9jYjWkEG7sxjWBjN6YRhr79E5PJRGPBY9B9zTNlqzmIJHMeJaQoEZEFGFUWmYXFV199tRrDwS9KROR5q/p/PB+1rqrCDN+vun/OKlNbVHGZaHV9vldVcScjfCqBkBm0dHNmaynuU3POBGsx/Oy9/ZMxxsZuTCvY2I1phKEH1bCfyj6H8q3Yj81st6vOw9fKBKMo35vno/wv5UfznNRxvD4HDx6sxrBvOWjiBfvMal2VH833xsExQO1/vvLKK33HqFLjGzdu7LS5Iq66vtJCeO3VdtHqOGbVqlVVHyf57Nixoxrz3HPPddpcbQio36uMXjQZ/GU3phFs7MY0go3dmEawsRvTCEOvVKOqofSDhQolUnCfKhXM51GZcZntlzirKlMSGqhFMhXoklmfTOUeFhaVYHno0KFOW2WLnXvuuVVfRvzLiIg8RgmWP/3pTzttdR8s9GUy0/jeAV1K+z3veU+nffbZZ1djPvvZz054LQC47777Ou1vfvOb1Zjdu3d32oNkvU1UotpfdmMawcZuTCPY2I1pBBu7MY0wVIFuZGRElgI+GiUScWSTEmAy+6pzhFimnJQSjVhEy+wHp+aozs1lmtUebZyZpzLjeL8zJWzx9ZUgtHbt2qqP11ZFnl111VWd9llnnVWN4b6dO3dWYzhiTkXr8X2oez3nnHM67UsuuaQa8/nPf77q4zLV3/3ud6sxXDZblem68sorO21VfvuOO+7otNVz5XeYBeSJSmL5y25MI9jYjWkEG7sxjTD0rDf2NzP+Fmcoqf23OZggUy1E+fWZQAaeo5qz8p3Yv1KBJkuWLOm0VfAHZ1Xt37+/GsPHqQAe7lNZZ8qPnj9/fqd9xhlnVGM4+ESVCOc12rRpUzWGNR51Hn4/Pv3pT1djrr766gnnB+g14vdVvR9f/epXO+09e/ZUY1hrUNdfvnx5p83bY6m+M888s9OeKCvOX3ZjGsHGbkwjpI09ImZFxKaIuLfXXhURGyNie0R8PyLqn5uNMdOGyXzZvwhg61HtrwP4RinlPAD7Adx6PCdmjDm+pAS6iFgG4HcA/E8AfxKjytJ1AMbSfdYD+O8A/mqi86j92VkUUQIIj8nsv6bKOfEYlSHEIp4S0XiMCirJ7PelShNxQIbKxFJ7rTO8RiqYiYVOtfYqQIQFuUzWm7o+i4hbtmypxnD2nnpm119/faf9/ve/vxrD93rPPfdUY9T19+7d22mrMl0s4qmy2SxqqveTxUf1fvC6btu2rdNWWZpjZL/sdwD4MwBjK30WgAOllDHL3QlgafJcxpgpoK+xR8TvAthbSnl4kAtExG0R8VBEPJQp1G+MOTFkfoz/IICbIuJjAE4BcDqAvwQwPyJO7n3dlwGofykIoJRyJ4A7AWDevHmDbblhjDlm+hp7KeUrAL4CABFxLYD/XEr5/Yj4ewCfBHAXgFsA3J04VxWAwv6O8n85ACGz/7ZCBdow7LeqY9gfz1SzAeoACFWmmtdHVa7hOaqkGw7i4IANoE5yUQk9al37JTMB9bqpwB/2N9Xe63yvF198cTVm9erVfa+1devWTlsFCylfm5+tSmBhlM7D73kmCYv9fAC46aabOm0O4PnOd74z7ryO5ffsX8aoWLcdoz78t47hXMaYE8ykwmVLKf8G4N96f/4FgPcd/ykZY04EjqAzphFs7MY0wpTvz86oPcE4QEQJFyx4KGGNs6MyvwpUgSYstigxLrPXG59HkdnHWwXwcJWTF154oRrD4p86j1rrhQsXdtqqegwLW+q5cpnoxYsX973+FVdcUY3h4BN1Hypghrnwwgurvu3bt3faSpzkKjiZ/flU5iaL0+rZX3TRRZ02B10pkXUMf9mNaQQbuzGNYGM3phGGXqmGfR72MZSvy/6eGsN+swpG4eNUkgf7+qp6SabijaqownNSCSSZgB0OGlFBJJywoRI4uOqJ8iNVddkPfOADnbbytXlNHnzwwWrMI4880mm/+93vrsZ8/OMf77TZPwbqNVPPnvdHV3oNBz0B9fZXquIrVwBW68F6EQcUAfWzVtd64IEHOm0OKDoeiTDGmBmOjd2YRrCxG9MINnZjGmGoAt3cuXPxrne9q9PHwQ6qnDELUCrYYNGiRZ22Er8yZas5QEQFaHAQh8pey5Skzuyrru6Dgy9UEEdmWyC+1+w2VqpUMsNrfe+991ZjWDT7whe+UI3hLDcl0PEaPf3009UYftYqyOjFF1+s+vh6LIgB9TNT51m5cmWnrcpvc9afeof4WW/evLnTVuLkGP6yG9MINnZjGsHGbkwjDNVnnz17dhVwwEEb7McBtS+pkio4WEf5RBwMo/xYrhaiklzY/1MBK8p34iAedW4ew8EYQJ0cwjoIkKvKyvehNIRMlVylK2zYsKHTfv7556sxX/rSlzrtz33uc33nqKrA8DbXKjFn1apVnTZvxQzoqr289TRvzwXUwUhqGyu+/xUrVlRjOKhIaQ+sX2W25h7DX3ZjGsHGbkwj2NiNaQQbuzGNMFSBbmRkpBLXOBhGlerlbYJUyV8WM3bv3l2NYaFPZbSxsKaEvkzZahWgwtdTY1hcUhlULL6pwBcO/FFzZmFNCW1K8OEKO0rs2rhxY6etAkRY7FLBOpw9qJ79s88+23cMV6FZs2ZNNUZlvXFQz7Jly6oxmdLaLH6qDEMWXlXlHH5GbBsTziE90hgzo7GxG9MINnZjGmHKq8uy/6d8ZPbruQ3UQRNcPQSoK4WqMZxkoqqFcDCMCrRQ/jj70So4iINGVDUdHqP8avaRVTAKBwOpMWo7LvYTf/7zn1djeHsl5Vt+73vf67TVs3/HO97Raat1zWyzzc9IaRjqfeBnpCoQsT+e2QpcbcXNlYNUheT3vve9nTb7+SrB5615jvs3xpjfKmzsxjSCjd2YRrCxG9MIoQIyTtjFIl4CsAPAQgD7+gyfbszEOQMzc96e8+CsKKWcrf5iqMb+1kUjHiqlXD70Cx8DM3HOwMyct+d8YvCP8cY0go3dmEaYKmO/c4queyzMxDkDM3PenvMJYEp8dmPM8PGP8cY0wtCNPSJujIgnI2J7RNw+7OtniIhvR8TeiHjsqL4FEXFfRDzV+3+d/DyFRMTyiNgQEU9ExOMR8cVe/7Sdd0ScEhEPRsSW3pz/vNe/KiI29t6R70dEHSQ+xUTErIjYFBH39trTfs5DNfaImAXgfwP4KIALAXwmIuoM/annrwHcSH23A7i/lLIWwP299nTiCIA/LaVcCOBKAH/YW9vpPO83AFxXSrkEwKUAboyIKwF8HcA3SinnAdgP4Napm+K4fBHA1qPa037Ow/6yvw/A9lLKL0opvwFwF4CbhzyHvpRS/h0Ap6TdDGB978/rAXximHPqRylldynlkd6fX8Poi7gU03jeZZSxdLjZvf8KgOsA/EOvf1rNGQAiYhmA3wHwf3vtwDSfMzB8Y18K4OgC2jt7fTOBxaWUsVpXewDU9aKmCRGxEsBlADZims+79+PwZgB7AdwH4GkAB0opY5vFTcd35A4AfwZgLJf1LEz/OVugG4Qy+iuMaflrjIiYB+AfAfxxKaWTjD8d511KGSmlXApgGUZ/8jt/amc0MRHxuwD2llIenuq5TJZhF6/YBWD5Ue1lvb6ZwIsRsaSUsjsilmD0SzStiIjZGDX0vyml/FOve9rPGwBKKQciYgOAqwDMj4iTe1/K6faOfBDATRHxMQCnADgdwF9ies8ZwPC/7D8DsLanXM4B8HsA7hnyHAblHgC39P58C4C7p3AuFT2/8VsAtpZS/uKov5q2846IsyNifu/PpwL4CEa1hg0APtkbNq3mXEr5SillWSllJUbf3/9XSvl9TOM5v0UpZaj/AfgYgG0Y9c3+67Cvn5zj3wLYDeAwRv2vWzHql90P4CkAPwawYKrnSXNeh9Ef0R8FsLn338em87wBvAfApt6cHwPw33r9qwE8CGA7gL8HMHeq5zrO/K8FcO9MmbMj6IxpBAt0xjSCjd2YRrCxG9MINnZjGsHGbkwj2NiNZCZkJ5rJ4V+9mYpeduI2jAa57MRoMNRnSilPTOnEzDHhL7tRzIjsRDM5bOxGMZOzE8042NiNaQQbu1HM5OxEMw42dqOYydmJZhyGnc9uZgCllCMR8UcA/hXALADfLqU8PsXTMseIf/VmTCP4x3hjGsHGbkwj2NiNaQQbuzGNYGM3phFs7MY0go3dmEawsRvTCP8fUfitLvj/1XgAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(img, cmap='gray')\n",
    "plt.xlabel(df.iloc[0]['emotion'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "bc947a81",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = []\n",
    "y = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "eadfee5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Creating a function to convert all the pixel data from string to int and appending it to X\n",
    "#And appending all the corresponding labels to y\n",
    "#Also dropping the disgust label for containing very less training data\n",
    "\n",
    "def getData(path):\n",
    "    anger = 0\n",
    "    fear = 0\n",
    "    sad = 0\n",
    "    happy = 0\n",
    "    surprise = 0\n",
    "    neutral = 0\n",
    "    df = pd.read_csv(path)\n",
    "    \n",
    "    X = []\n",
    "    y = []    \n",
    "    \n",
    "    for i in range(len(df)):\n",
    "        if df.iloc[i]['emotion'] != 1:\n",
    "            if df.iloc[i]['emotion'] == 0:\n",
    "                if anger <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    anger += 1\n",
    "                else:\n",
    "                    pass\n",
    "                \n",
    "            if df.iloc[i]['emotion'] == 2:\n",
    "                if fear <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    fear += 1\n",
    "                else:\n",
    "                    pass\n",
    "                \n",
    "            if df.iloc[i]['emotion'] == 3:\n",
    "                if happy <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    happy += 1\n",
    "                else:\n",
    "                    pass\n",
    "                \n",
    "            if df.iloc[i]['emotion'] == 4:\n",
    "                if sad <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    sad += 1\n",
    "                else:\n",
    "                    pass\n",
    "                \n",
    "            if df.iloc[i]['emotion'] == 5:\n",
    "                if surprise <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    surprise += 1\n",
    "                else:\n",
    "                    pass\n",
    "                \n",
    "            if df.iloc[i]['emotion'] == 6:\n",
    "                if neutral <= 4000:            \n",
    "                    y.append(df.iloc[i]['emotion'])\n",
    "                    im = df.iloc[i]['pixels']\n",
    "                    im = [int(x) for x in im.split()]\n",
    "                    X.append(im)\n",
    "                    neutral += 1\n",
    "                else:\n",
    "                    pass\n",
    "\n",
    "            \n",
    "            \n",
    "    return X, y "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9182caab",
   "metadata": {},
   "outputs": [],
   "source": [
    "X, y = getData('fer2013.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "2f1729db",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 2, 3, 4, 5, 6], dtype=int64),\n",
       " array([4001, 4001, 4001, 4001, 4001, 4001], dtype=int64))"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(y, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "77f7611c",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = np.array(X)/255.0\n",
    "y = np.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "fc9d364c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((24006, 2304), (24006,))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape, y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "7283aa01",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_o = []\n",
    "for i in y:\n",
    "    if i != 6:\n",
    "        y_o.append(i)\n",
    "        \n",
    "    else:\n",
    "        y_o.append(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "b937efa1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([0, 1, 2, 3, 4, 5], dtype=int64),\n",
       " array([4001, 4001, 4001, 4001, 4001, 4001], dtype=int64))"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.unique(y_o, return_counts=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "e94e8b9b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(5):\n",
    "    r = np.random.randint((1), 24000, 1)[0]\n",
    "    plt.figure()\n",
    "    plt.imshow(X[r].reshape(48,48), cmap='gray')\n",
    "    plt.xlabel(label_map[y_o[r]])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "939c575b",
   "metadata": {},
   "outputs": [],
   "source": [
    "X = X.reshape(len(X), 48, 48, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "a0f11e9e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24006, 48, 48, 1)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "7378aff6",
   "metadata": {},
   "outputs": [],
   "source": [
    "y_new = to_categorical(y_o, num_classes=6)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "7dcb33f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(24006, (24006, 6))"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(y_o), y_new.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "67f4a28a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, array([1., 0., 0., 0., 0., 0.], dtype=float32))"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_o[150], y_new[150]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8976ba94",
   "metadata": {},
   "source": [
    "### Creating the training model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "846b977d",
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "\n",
    "input_shape = (48,48,1)\n",
    "\n",
    "\n",
    "model.add(Conv2D(64, (5, 5), input_shape=input_shape,activation='relu', padding='same'))\n",
    "model.add(Conv2D(64, (5, 5), padding='same'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "\n",
    "model.add(Conv2D(128, (5, 5),activation='relu',padding='same'))\n",
    "model.add(Conv2D(128, (5, 5),padding='same'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(Activation('relu'))\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "model.add(Conv2D(256, (3, 3),activation='relu',padding='same'))\n",
    "model.add(Conv2D(256, (3, 3),activation='relu',padding='same'))\n",
    "model.add(BatchNormalization())\n",
    "model.add(MaxPooling2D(pool_size=(2, 2)))\n",
    "\n",
    "\n",
    "model.add(Flatten())\n",
    "model.add(Dense(6, activation='softmax'))\n",
    "\n",
    "model.compile(loss='categorical_crossentropy', metrics=['accuracy'],optimizer='adam')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "20658eb5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/22\n",
      "301/301 [==============================] - 888s 3s/step - loss: 2.1145 - accuracy: 0.2491 - val_loss: 2.3282 - val_accuracy: 0.1125\n",
      "Epoch 2/22\n",
      "301/301 [==============================] - 875s 3s/step - loss: 1.6994 - accuracy: 0.3391 - val_loss: 1.4601 - val_accuracy: 0.4202\n",
      "Epoch 3/22\n",
      "301/301 [==============================] - 876s 3s/step - loss: 1.5329 - accuracy: 0.4071 - val_loss: 1.7057 - val_accuracy: 0.3344\n",
      "Epoch 4/22\n",
      "301/301 [==============================] - 882s 3s/step - loss: 1.4008 - accuracy: 0.4560 - val_loss: 1.9871 - val_accuracy: 0.3344\n",
      "Epoch 5/22\n",
      "301/301 [==============================] - 873s 3s/step - loss: 1.2976 - accuracy: 0.4979 - val_loss: 1.3370 - val_accuracy: 0.4621\n",
      "Epoch 6/22\n",
      "301/301 [==============================] - 822s 3s/step - loss: 1.2430 - accuracy: 0.5204 - val_loss: 1.3110 - val_accuracy: 0.4971\n",
      "Epoch 7/22\n",
      "301/301 [==============================] - 823s 3s/step - loss: 1.1762 - accuracy: 0.5494 - val_loss: 1.5014 - val_accuracy: 0.4157\n",
      "Epoch 8/22\n",
      "301/301 [==============================] - 823s 3s/step - loss: 1.0975 - accuracy: 0.5721 - val_loss: 1.9426 - val_accuracy: 0.3898\n",
      "Epoch 9/22\n",
      "301/301 [==============================] - 827s 3s/step - loss: 1.0387 - accuracy: 0.6054 - val_loss: 1.2653 - val_accuracy: 0.5531\n",
      "Epoch 10/22\n",
      "301/301 [==============================] - 829s 3s/step - loss: 0.9660 - accuracy: 0.6339 - val_loss: 1.5575 - val_accuracy: 0.4550\n",
      "Epoch 11/22\n",
      "301/301 [==============================] - 824s 3s/step - loss: 0.9016 - accuracy: 0.6594 - val_loss: 1.3012 - val_accuracy: 0.5277\n",
      "Epoch 12/22\n",
      "301/301 [==============================] - 825s 3s/step - loss: 0.8054 - accuracy: 0.7005 - val_loss: 1.4550 - val_accuracy: 0.5165\n",
      "Epoch 13/22\n",
      "301/301 [==============================] - 825s 3s/step - loss: 0.7042 - accuracy: 0.7373 - val_loss: 2.0109 - val_accuracy: 0.4150\n",
      "Epoch 14/22\n",
      "301/301 [==============================] - 824s 3s/step - loss: 0.6147 - accuracy: 0.7730 - val_loss: 2.0443 - val_accuracy: 0.4484\n",
      "Epoch 15/22\n",
      "301/301 [==============================] - 822s 3s/step - loss: 0.4691 - accuracy: 0.8306 - val_loss: 1.7335 - val_accuracy: 0.5219\n",
      "Epoch 16/22\n",
      "301/301 [==============================] - 838s 3s/step - loss: 0.3539 - accuracy: 0.8737 - val_loss: 2.1641 - val_accuracy: 0.4819\n",
      "Epoch 17/22\n",
      "301/301 [==============================] - 823s 3s/step - loss: 0.2974 - accuracy: 0.8944 - val_loss: 2.5797 - val_accuracy: 0.4600\n",
      "Epoch 18/22\n",
      "301/301 [==============================] - 824s 3s/step - loss: 0.2162 - accuracy: 0.9231 - val_loss: 2.4675 - val_accuracy: 0.5104\n",
      "Epoch 19/22\n",
      "301/301 [==============================] - 823s 3s/step - loss: 0.2022 - accuracy: 0.9344 - val_loss: 2.3436 - val_accuracy: 0.5254\n",
      "Epoch 20/22\n",
      "301/301 [==============================] - 827s 3s/step - loss: 0.1193 - accuracy: 0.9668 - val_loss: 2.6665 - val_accuracy: 0.5158\n",
      "Epoch 21/22\n",
      "301/301 [==============================] - 824s 3s/step - loss: 0.1241 - accuracy: 0.9621 - val_loss: 3.0195 - val_accuracy: 0.5319\n",
      "Epoch 22/22\n",
      "301/301 [==============================] - 824s 3s/step - loss: 0.1339 - accuracy: 0.9574 - val_loss: 3.6501 - val_accuracy: 0.4327\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1c3a17d2be0>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(X, y_new, epochs=22, batch_size=64, shuffle=True, validation_split=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "6e965df4",
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save('model.h5')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e64ae18d",
   "metadata": {},
   "source": [
    "### Creating the app"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "bbdc0661",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overwriting app.py\n"
     ]
    }
   ],
   "source": [
    "%%writefile app.py\n",
    "import streamlit as st\n",
    "import tensorflow as tf\n",
    "\n",
    "#st.set_option('deprecation.showFileUploaderEncoding', False)\n",
    "\n",
    "@st.cache(allow_output_mutation=True)\n",
    "def load_model():\n",
    "  model=tf.keras.models.load_model('C:\\\\Users\\\\deshr\\\\OneDrive\\\\Desktop\\\\Elevate AI Take Home\\\\model.h5')\n",
    "  return model\n",
    "with st.spinner('Model is being loaded..'):\n",
    "  model=load_model()\n",
    "\n",
    "st.write(\"\"\"\n",
    "         # Facial Expression Recognition\n",
    "         \"\"\"\n",
    "         )\n",
    "\n",
    "file = st.file_uploader(\"Please upload a picture of your face..\", type=[\"jpg\", \"png\"])\n",
    "import cv2\n",
    "from PIL import Image, ImageOps\n",
    "import numpy as np\n",
    "st.set_option('deprecation.showfileUploaderEncoding', False)\n",
    "def import_and_predict(image_data, model):\n",
    "    \n",
    "        size = (48,48)    \n",
    "        image = ImageOps.fit(image_data, size, Image.ANTIALIAS)\n",
    "        image = np.asarray(image)\n",
    "        img = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)\n",
    "        img_resize = (cv2.resize(img, dsize=(48, 48)))\n",
    "        \n",
    "        img_reshape = img.reshape(1,48,48,1)\n",
    "    \n",
    "        prediction = model.predict(img_reshape)\n",
    "        \n",
    "        return prediction\n",
    "\n",
    "if file is None:\n",
    "    st.text(\"Please upload an image file\")\n",
    "else:\n",
    "    image = Image.open(file)\n",
    "    st.image(image, use_column_width=True)\n",
    "    predictions = import_and_predict(image, model)\n",
    "    class_names = ['Angry', 'Neutral', 'Scared', 'Happy', 'Sad', 'Surprised']\n",
    "    score = tf.nn.softmax(predictions[0])\n",
    "    #st.write(predictions)\n",
    "    st.write(\"The person in the image is {} \".format(class_names[np.argmax(predictions)]))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35afd40d",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}