testing_models.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "animal-audit",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "import random\n",
    "import torchvision\n",
    "import numpy as np\n",
    "from torch.utils.data.dataset import Dataset\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import transforms\n",
    "from PIL import Image\n",
    "import copy\n",
    "import os\n",
    "import traceback\n",
    "import datetime\n",
    "from torchvision import datasets, transforms, models\n",
    "from efficientnet_pytorch import EfficientNet\n",
    "import math\n",
    "import time\n",
    "import matplotlib.pyplot as plt\n",
    "from IPython.display import clear_output\n",
    "import cv2\n",
    "import pandas as pd\n",
    "from random import shuffle\n",
    "from sklearn.metrics import roc_curve, roc_auc_score, auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "id": "durable-score",
   "metadata": {},
   "outputs": [],
   "source": [
    "male_test_dir = 'TestDataset/CroppedFaces/1/'\n",
    "female_test_dir = 'TestDataset/CroppedFaces/0/'\n",
    "male_widerfaces_dir = 'WiderFaces/CroppedFaces/1/'\n",
    "female_widerfaces_dir = 'WiderFaces/CroppedFaces/0/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "israeli-ireland",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "distinct-beads",
   "metadata": {},
   "source": [
    "# EfficientNet trained on WiderFaces"
   ]
  },
  {
   "attachments": {
    "%D0%B8%D0%B7%D0%BE%D0%B1%D1%80%D0%B0%D0%B6%D0%B5%D0%BD%D0%B8%D0%B5-3.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "classical-exemption",
   "metadata": {},
   "source": [
    "![%D0%B8%D0%B7%D0%BE%D0%B1%D1%80%D0%B0%D0%B6%D0%B5%D0%BD%D0%B8%D0%B5-3.png](attachment:%D0%B8%D0%B7%D0%BE%D0%B1%D1%80%D0%B0%D0%B6%D0%B5%D0%BD%D0%B8%D0%B5-3.png)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 101,
   "id": "threatened-fields",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "train_loss_b4 = np.array([0.59985099, 0.56365057, 0.55195486, 0.54679735, 0.53908001,\n",
    "       0.5342011 , 0.53069071, 0.52794463, 0.52921178, 0.52971087,\n",
    "       0.525452  , 0.5253335 , 0.52518364, 0.52081931, 0.52084775,\n",
    "       0.52326544, 0.52185264, 0.5234638 , 0.52026343, 0.52350796,\n",
    "       0.52316783, 0.52188409, 0.5198656 , 0.51939222, 0.51874303,\n",
    "       0.52051153, 0.52000468, 0.51827887, 0.51800349, 0.51918827,\n",
    "       0.51930214, 0.51656442, 0.51645   , 0.51684091, 0.51794271,\n",
    "       0.51616254, 0.51703963, 0.51948734, 0.51900152, 0.51573063,\n",
    "       0.51642533, 0.51777791, 0.51816169, 0.52035771, 0.51714839,\n",
    "       0.51864058])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "visible-facial",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "train_loss_b2 = np.array([0.6052186 , 0.56911385, 0.5574453 , 0.55016615, 0.54168102,\n",
    "       0.53894525, 0.53446001, 0.53152139, 0.53218537, 0.53115029,\n",
    "       0.52659135, 0.52361184, 0.52186587, 0.52220552, 0.52411116,\n",
    "       0.52279879, 0.52124053, 0.52083742, 0.52303386, 0.52011177,\n",
    "       0.52033576, 0.5185603 , 0.52080499, 0.52124911, 0.52336861,\n",
    "       0.52385182, 0.51835908, 0.51742606, 0.51776006, 0.52208002,\n",
    "       0.5215389 , 0.51827229, 0.5181963 , 0.519534  , 0.51791218,\n",
    "       0.5208478 , 0.51998415, 0.51815542, 0.51819823, 0.51872568,\n",
    "       0.51862968, 0.51707046, 0.51623882, 0.51550393, 0.5155136 ,\n",
    "       0.51570724, 0.51681575, 0.51560851, 0.51775145, 0.52019937,\n",
    "       0.51687286, 0.51716193, 0.51615627, 0.51437265, 0.5131944 ,\n",
    "       0.51548408, 0.51490865, 0.51211991, 0.51397634, 0.51724344,\n",
    "       0.51672456, 0.51425292, 0.51556877, 0.51441405, 0.51428412,\n",
    "       0.51329801, 0.51713841, 0.51626742, 0.51434804, 0.51466166,\n",
    "       0.51629464, 0.51677432, 0.51557726, 0.51343675, 0.51304958,\n",
    "       0.5150763 , 0.51453029])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "id": "defensive-staff",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "train_loss_b0 = np.array([0.60856076, 0.57111408, 0.56099927, 0.55100713, 0.54825293,\n",
    "       0.54394683, 0.53996492, 0.5354263 , 0.53334227, 0.53069973,\n",
    "       0.52839631, 0.52738957, 0.52686879, 0.5241123 , 0.52378575,\n",
    "       0.52310428, 0.52252396, 0.52261295, 0.52228643, 0.52185731,\n",
    "       0.52134645, 0.51845763, 0.51941696, 0.51881798, 0.52049083,\n",
    "       0.5188519 , 0.51859298, 0.52013841, 0.52021711, 0.51773428,\n",
    "       0.51545191, 0.51574762, 0.51617709, 0.51644041, 0.5157785 ,\n",
    "       0.51686641, 0.51540858, 0.51500336, 0.51501249, 0.51628815,\n",
    "       0.51571677, 0.51459266, 0.51391325, 0.51446047, 0.51389925,\n",
    "       0.51506632, 0.51373299, 0.51296119, 0.51623298, 0.51637701,\n",
    "       0.51739708, 0.51638749, 0.51580694, 0.51584595, 0.51818495,\n",
    "       0.51472711, 0.51504064, 0.51506162, 0.51875502, 0.5135307 ,\n",
    "       0.51279921, 0.51405837, 0.5130916 , 0.51420885, 0.51471513,\n",
    "       0.51377036, 0.51513551, 0.51481467, 0.51644905, 0.51559136,\n",
    "       0.51441392, 0.51687581, 0.51259916, 0.51224624, 0.51268969,\n",
    "       0.51415301, 0.51350467, 0.51389328, 0.51517838, 0.51422734,\n",
    "       0.51297072, 0.51421655, 0.51216522, 0.51387969, 0.51533993,\n",
    "       0.51430353, 0.51533056, 0.51539622, 0.51447393, 0.51307328,\n",
    "       0.51401379, 0.51558578, 0.51619216, 0.5127755 , 0.51366809,\n",
    "       0.51281459, 0.51342048, 0.51236066, 0.51167401, 0.51235478,\n",
    "       0.51243799, 0.51199771, 0.51168784, 0.51292284, 0.51207089,\n",
    "       0.5139119 , 0.51293784, 0.51303459, 0.5124369 , 0.51729528,\n",
    "       0.51236272, 0.51111084, 0.51069446, 0.51310922])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "id": "hungry-bride",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "train_loss_resnet_50 = np.array([0.64653235, 0.6203505 , 0.61718526, 0.61393385, 0.61307396,\n",
    "       0.61182842, 0.61240318, 0.61031286, 0.6098428 , 0.60857303,\n",
    "       0.60850796, 0.60867264, 0.60902749, 0.60466904, 0.6079799 ,\n",
    "       0.607366  , 0.60490584, 0.60805841, 0.60660642, 0.6055448 ,\n",
    "       0.60535359, 0.60741585, 0.60462402, 0.6041952 , 0.60272666,\n",
    "       0.60244788, 0.60303182, 0.60419609, 0.60242472, 0.60225578,\n",
    "       0.60301199, 0.60206481, 0.60140062, 0.59988652, 0.59918293,\n",
    "       0.60084441, 0.59798817, 0.59897382, 0.599241  , 0.59663436,\n",
    "       0.59781739, 0.5958732 , 0.59572171, 0.59482191, 0.59500719,\n",
    "       0.5978633 , 0.59482827, 0.59347686, 0.5924632 , 0.59191896,\n",
    "       0.59385116, 0.58978909, 0.59220737, 0.59203889, 0.59018709,\n",
    "       0.5877888 , 0.59091345, 0.58771855, 0.58886294, 0.58689065,\n",
    "       0.58511811, 0.58576459, 0.58548203, 0.58721575, 0.58681743,\n",
    "       0.58577648, 0.58473845, 0.58403157, 0.58478771, 0.58363727,\n",
    "       0.58310269, 0.58100876, 0.58186875])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "id": "elect-ireland",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def draw_trainloss():\n",
    "    fig = plt.figure()\n",
    "    fig.set_size_inches(9, 4)\n",
    "\n",
    "    min_loss = min(np.concatenate((train_loss_b4, train_loss_b2, train_loss_b0, train_loss_resnet_50)))\n",
    "    #plt.axline((0,min_loss), (len(train_loss_b4),min_loss))\n",
    "    plt.plot(range(0, len(train_loss_b4)),train_loss_b4, label='b4')\n",
    "    plt.plot(range(0, len(train_loss_b2)),train_loss_b2, label='b2')\n",
    "    plt.plot(range(0, len(train_loss_b0)),train_loss_b0, label='b0')\n",
    "    #plt.plot(range(0, len(train_loss_resnet_50)),train_loss_resnet_50, label='resnet')\n",
    "    plt.legend()\n",
    "    plt.grid(True)\n",
    "    plt.xlabel('epochs', fontsize=18)\n",
    "    plt.ylabel('train loss', fontsize=16)\n",
    "draw_trainloss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "reflected-douglas",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "f1_b4_more_epoch = np.array([0.77234825, 0.85938961, 0.88914612, 0.90028133, 0.92052184,\n",
    "       0.93542485, 0.93772361, 0.94581791, 0.94291558, 0.94267163,\n",
    "       0.95253327, 0.9521066 , 0.95210683, 0.96359274, 0.96407998,\n",
    "       0.95581637, 0.96173501, 0.95585346, 0.9649098 , 0.95628248,\n",
    "       0.95670557, 0.95931586, 0.96537666, 0.9649179 , 0.96834817,\n",
    "       0.96228779, 0.96415561, 0.97018646, 0.96803818, 0.9657293 ,\n",
    "       0.96649631, 0.97184706, 0.97163587, 0.97275875, 0.97001996,\n",
    "       0.97482568, 0.97187484, 0.96504725, 0.96747651, 0.97451388,\n",
    "       0.97274471, 0.97104397, 0.9677098 , 0.96330232, 0.97124903,\n",
    "       0.96994772])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "impressed-vinyl",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "f1_b0 = np.array([0.74553584, 0.84417013, 0.86901644, 0.89541713, 0.89675685,\n",
    "       0.90837259, 0.91768327, 0.92634407, 0.93334793, 0.93861419,\n",
    "       0.94485894, 0.94655801, 0.94916649, 0.95617493, 0.95445069,\n",
    "       0.95658784, 0.95878777, 0.95785487, 0.95943976, 0.96222041,\n",
    "       0.96219046, 0.96698898, 0.96636926, 0.96855329, 0.96508292,\n",
    "       0.96673766, 0.9685896 , 0.96372741, 0.96355993, 0.96858601,\n",
    "       0.97458905, 0.97481998, 0.97297571, 0.97232471, 0.97274618,\n",
    "       0.97237567, 0.976833  , 0.97703221, 0.97615728, 0.97325634,\n",
    "       0.97410607, 0.97756957, 0.9799869 , 0.9780846 , 0.97902839,\n",
    "       0.97659952, 0.97889293, 0.98050815, 0.97359794, 0.97349262,\n",
    "       0.97159137, 0.97328786, 0.97337079, 0.9760215 , 0.96954808,\n",
    "       0.97755936, 0.97534928, 0.97690441, 0.97005164, 0.97907219,\n",
    "       0.981873  , 0.97907422, 0.980598  , 0.97880763, 0.97807784,\n",
    "       0.9785275 , 0.97552325, 0.97677667, 0.97257429, 0.97469843,\n",
    "       0.97826804, 0.97315162, 0.98303544, 0.98241418, 0.9819695 ,\n",
    "       0.97877178, 0.98135277, 0.98030618, 0.97650908, 0.97942035,\n",
    "       0.98202513, 0.97807217, 0.98337719, 0.9786321 , 0.97544576,\n",
    "       0.97860466, 0.97777321, 0.97710191, 0.97701971, 0.98171948,\n",
    "       0.97930736, 0.97490165, 0.97325395, 0.98252318, 0.9788002 ,\n",
    "       0.98258004, 0.98151398, 0.98345829, 0.9834708 , 0.98229496,\n",
    "       0.9823409 , 0.98360852, 0.98448353, 0.98209729, 0.98259339,\n",
    "       0.97781158, 0.98132296, 0.98118523, 0.98278916, 0.97040269,\n",
    "       0.98126554, 0.98543539, 0.98576891, 0.98159647])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "level-watershed",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "resnet_50_f1 = np.array([0.51717312, 0.70026124, 0.70427031, 0.71696337, 0.72075081,\n",
    "       0.72430848, 0.72565674, 0.73229351, 0.7314867 , 0.73884242,\n",
    "       0.73105033, 0.74530534, 0.73810617, 0.75299731, 0.73917709,\n",
    "       0.7367179 , 0.75003489, 0.73559505, 0.74430069, 0.74487025,\n",
    "       0.74826724, 0.73252256, 0.74855785, 0.74808751, 0.76021572,\n",
    "       0.75643408, 0.76173415, 0.74940503, 0.75284194, 0.76377917,\n",
    "       0.75198575, 0.75605644, 0.76123852, 0.76048536, 0.76017506,\n",
    "       0.76123278, 0.77529628, 0.76634178, 0.76838224, 0.77133836,\n",
    "       0.76738398, 0.77703093, 0.77775145, 0.77535304, 0.77924553,\n",
    "       0.77245922, 0.78082615, 0.77879033, 0.78717026, 0.78609693,\n",
    "       0.77751601, 0.79018782, 0.79049522, 0.78437521, 0.79236709,\n",
    "       0.79738984, 0.78992533, 0.80163305, 0.79780963, 0.80270017,\n",
    "       0.80481374, 0.80367896, 0.80251467, 0.80085923, 0.79763415,\n",
    "       0.80295778, 0.80918511, 0.81147141, 0.80700816, 0.80831296,\n",
    "       0.81155167, 0.8167119 , 0.81617485])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ahead-study",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "nervous-strap",
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "from efficientnet_pytorch import EfficientNet\n",
    "class MyModel(nn.Module):\n",
    "    def __init__(self, model_name: str):\n",
    "        super(MyModel,self).__init__()\n",
    "\n",
    "        # self.model = torch.hub.load('pytorch/vision:v0.9.0', 'inception_v3', pretrained=True)\n",
    "        self.model = EfficientNet.from_pretrained(model_name)\n",
    "        self.model.set_swish(memory_efficient=False)\n",
    "        self.linear1 = torch.nn.Linear(1000, 1)\n",
    "        self.activation3 = torch.nn.Sigmoid()\n",
    "\n",
    "    def forward(self,x):\n",
    "        answ = self.model(x)\n",
    "        #print(answ.shape)\n",
    "        #answ = self.encoder(answ)\n",
    "        #print(answ.shape)\n",
    "        answ = self.activation3(self.linear1(answ))\n",
    "\n",
    "        return answ\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "grateful-respect",
   "metadata": {},
   "outputs": [],
   "source": [
    "transform_efnet=transforms.Compose([transforms.Resize((150,150)),\n",
    "                              transforms.ToTensor()\n",
    "                              ])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "complimentary-cocktail",
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def init_ef_model(model_name, model_path):\n",
    "    device = torch.device('cpu')\n",
    "    model = MyModel(model_name)\n",
    "    model.load_state_dict(torch.load(model_path, map_location=device))\n",
    "    model.eval();\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "superb-roots",
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def get_res_efnet(image_path: str, model: MyModel):\n",
    "    image = Image.open(image_path)\n",
    "    image = transform_efnet(image).unsqueeze(0)\n",
    "\n",
    "    # model.cuda()\n",
    "    # Set layers such as dropout and batchnorm in evaluation mode\n",
    "    # model.eval();\n",
    "\n",
    "    # Get the 1000-dimensional model output\n",
    "    with torch.no_grad():\n",
    "        out = model(image)\n",
    "    # print(out.detach().numpy()[0][0])    \n",
    "    return out.detach().numpy()[0][0]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "sporting-devon",
   "metadata": {},
   "source": [
    "## EfficientNet-b0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "id": "micro-uncertainty",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_b0_name = 'efficientnet-b0'\n",
    "model_b0_path = 'PretrainedModels/GenderEstimation/Ours/ef_b0.pth'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "id": "adjustable-classic",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded pretrained weights for efficientnet-b0\n"
     ]
    }
   ],
   "source": [
    "b0_model = init_ef_model(model_b0_name, model_b0_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "id": "legislative-ethernet",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_res_b0(image_path: str):\n",
    "    return get_res_efnet(image_path, b0_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "centered-eating",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "id": "opposite-rebel",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 170, FP: 14, TN: 100, FN: 25\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.8737864077669902\n",
      "Precision: 0.9239130434782609\n",
      "Recall: 0.8717948717948718\n",
      "F1_Score: 0.8970976253298153\n"
     ]
    }
   ],
   "source": [
    "acc_b0, prec_b0, rec_b0, f1_b0 = get_scores(get_res_b0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "separated-peninsula",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "missing-distributor",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 485, FP: 19, TN: 481, FN: 15\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.966\n",
      "Precision: 0.9623015873015873\n",
      "Recall: 0.97\n",
      "F1_Score: 0.9661354581673307\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.966, 0.9623015873015873, 0.97, 0.9661354581673307)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_b0, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "joined-joining",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "id": "lesbian-notice",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 0.2373650074005127\n",
      "Max time: 0.3919520378112793\n",
      "Avg time: 0.27817320251464844\n"
     ]
    }
   ],
   "source": [
    "min_time_b0, max_time_b0, avg_time_b0 = get_times(get_res_b0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "twenty-sweden",
   "metadata": {},
   "source": [
    "## EfficientNet-b2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "id": "introductory-example",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_b2_name = 'efficientnet-b2'\n",
    "model_b2_path = 'PretrainedModels/GenderEstimation/Ours/ef_b2_more_epochs.pth'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "id": "electoral-france",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded pretrained weights for efficientnet-b2\n"
     ]
    }
   ],
   "source": [
    "b2_model = init_ef_model(model_b2_name, model_b2_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "id": "coated-roommate",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_res_b2(image_path: str):\n",
    "    return get_res_efnet(image_path, b2_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "common-happiness",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "id": "valid-vision",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 173, FP: 17, TN: 97, FN: 22\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.8737864077669902\n",
      "Precision: 0.9105263157894737\n",
      "Recall: 0.8871794871794871\n",
      "F1_Score: 0.8987012987012988\n"
     ]
    }
   ],
   "source": [
    "acc_b2, prec_b2, rec_b2, f1_b2 = get_scores(get_res_b2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "flush-panel",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "advisory-holiday",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 485, FP: 11, TN: 489, FN: 15\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.974\n",
      "Precision: 0.9778225806451613\n",
      "Recall: 0.97\n",
      "F1_Score: 0.9738955823293172\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.974, 0.9778225806451613, 0.97, 0.9738955823293172)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_b2, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "surprising-response",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "id": "after-surrey",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 0.3919541835784912\n",
      "Max time: 1.1987965106964111\n",
      "Avg time: 0.5123051686286926\n"
     ]
    }
   ],
   "source": [
    "min_time_b2, max_time_b2, avg_time_b2 = get_times(get_res_b2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "vocational-queensland",
   "metadata": {},
   "source": [
    "## EfficientNet-b4"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "id": "alive-malpractice",
   "metadata": {},
   "outputs": [],
   "source": [
    "model_b4_name = 'efficientnet-b4'\n",
    "model_b4_path = 'PretrainedModels/GenderEstimation/Ours/ef_b4.pth'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 171,
   "id": "sporting-diploma",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loaded pretrained weights for efficientnet-b4\n"
     ]
    }
   ],
   "source": [
    "b4_model = init_ef_model(model_b4_name, model_b4_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "id": "gentle-charlotte",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_res_b4(image_path: str):\n",
    "    return get_res_efnet(image_path, b4_model)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "proprietary-qatar",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "forbidden-imperial",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 175, FP: 19, TN: 95, FN: 20\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.8737864077669902\n",
      "Precision: 0.9020618556701031\n",
      "Recall: 0.8974358974358975\n",
      "F1_Score: 0.8997429305912596\n"
     ]
    }
   ],
   "source": [
    "acc_b4, prec_b4, rec_b4, f1_b4 = get_scores(get_res_b4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fantastic-contract",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "spanish-education",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 487, FP: 13, TN: 487, FN: 13\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.974\n",
      "Precision: 0.974\n",
      "Recall: 0.974\n",
      "F1_Score: 0.974\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.974, 0.974, 0.974, 0.974)"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_b4, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "familiar-efficiency",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "id": "underlying-outdoors",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 0.7619607448577881\n",
      "Max time: 1.6595890522003174\n",
      "Avg time: 0.8659273972511291\n"
     ]
    }
   ],
   "source": [
    "min_time_b4, max_time_b4, avg_time_b4 = get_times(get_res_b4)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "modular-fancy",
   "metadata": {},
   "source": [
    "# Gil Levi and Tal Hassner model trained on Adience"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "id": "downtown-place",
   "metadata": {},
   "outputs": [],
   "source": [
    "genderProto = \"PretrainedModels/GenderEstimation/OthersModels/gender_deploy.prototxt\"\n",
    "genderModel = \"PretrainedModels/GenderEstimation/OthersModels/gender_net.caffemodel\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "id": "contained-franklin",
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def get_res_adience(image_path):\n",
    "    genderNet = cv2.dnn.readNet(genderModel, genderProto)\n",
    "    genderList = ['Male', 'Female']\n",
    "    face = cv2.imread(image_path)\n",
    "    MODEL_MEAN_VALUES = (78.4263377603, 87.7689143744, 114.895847746)\n",
    "    blob = cv2.dnn.blobFromImage(face, 1, (227, 227), MODEL_MEAN_VALUES, swapRB=False)\n",
    "    genderNet.setInput(blob)\n",
    "    genderPreds = genderNet.forward()\n",
    "    gender = genderList[genderPreds[0].argmax()]\n",
    "    if gender == 'Female':\n",
    "        return 0\n",
    "    else:\n",
    "        return 1"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "northern-original",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "id": "national-savannah",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 117, FP: 14, TN: 100, FN: 78\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.7022653721682848\n",
      "Precision: 0.8931297709923665\n",
      "Recall: 0.6\n",
      "F1_Score: 0.7177914110429446\n"
     ]
    }
   ],
   "source": [
    "acc_adience, prec_adience, rec_adience, f1_adience = get_scores(get_res_adience)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "recreational-master",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "adult-hanging",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 366, FP: 125, TN: 375, FN: 132\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.7424849699398798\n",
      "Precision: 0.745417515274949\n",
      "Recall: 0.7349397590361446\n",
      "F1_Score: 0.7401415571284126\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.7424849699398798, 0.745417515274949, 0.7349397590361446, 0.7401415571284126)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_adience, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "broadband-longer",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 114,
   "id": "alien-species",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 0.11764931678771973\n",
      "Max time: 0.24235272407531738\n",
      "Avg time: 0.13374355888366699\n"
     ]
    }
   ],
   "source": [
    "min_time_adience, max_time_adience, avg_time_adience = get_times(get_res_adience)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "joined-hypothetical",
   "metadata": {},
   "source": [
    "# VGG-Face model trained on IMDb-Wiki"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 175,
   "id": "comparative-location",
   "metadata": {},
   "outputs": [],
   "source": [
    "from deepface import DeepFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 176,
   "id": "prostate-arrival",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_res_vgg_imdb(image_path):\n",
    "    obj = DeepFace.analyze(image_path, actions = ['gender'], enforce_detection=False)\n",
    "    gender = obj['gender']\n",
    "    res = 1 if gender == 'Man' else 0\n",
    "    clear_output()\n",
    "    return res"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bright-defeat",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 102,
   "id": "criminal-muscle",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 192, FP: 104, TN: 10, FN: 3\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.6537216828478964\n",
      "Precision: 0.6486486486486487\n",
      "Recall: 0.9846153846153847\n",
      "F1_Score: 0.7820773930753564\n"
     ]
    }
   ],
   "source": [
    "acc_vgg_imdb, prec_vgg_imdb, rec_vgg_imdb, f1_vgg_imdb = get_scores(get_res_vgg_imdb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "broad-assist",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "suffering-strain",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 485, FP: 350, TN: 150, FN: 15\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.635\n",
      "Precision: 0.5808383233532934\n",
      "Recall: 0.97\n",
      "F1_Score: 0.7265917602996255\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.635, 0.5808383233532934, 0.97, 0.7265917602996255)"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_vgg_imdb, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "knowing-opportunity",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 115,
   "id": "vital-emergency",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 1.568796157836914\n",
      "Max time: 2.8674840927124023\n",
      "Avg time: 1.7493990716934205\n"
     ]
    }
   ],
   "source": [
    "min_time_vgg_imdb, max_time_vgg_imdb, avg_time_vgg_imdb = get_times(get_res_vgg_imdb)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "documentary-fleet",
   "metadata": {},
   "source": [
    "# FairFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 177,
   "id": "precise-climb",
   "metadata": {},
   "outputs": [],
   "source": [
    "import dlib"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "id": "labeled-allocation",
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def get_res_fairface(image_path):\n",
    "    device = torch.device('cpu')\n",
    "    model_fair_4 = torchvision.models.resnet34(pretrained=True)\n",
    "    model_fair_4.fc = nn.Linear(model_fair_4.fc.in_features, 18)\n",
    "    model_fair_4.load_state_dict(torch.load('PretrainedModels/GenderEstimation/OthersModels/fairface_alldata_4race_20191111.pt', map_location=device))\n",
    "    model_fair_4 = model_fair_4.to(device)\n",
    "    model_fair_4.eval()\n",
    "    trans = transforms.Compose([\n",
    "        transforms.ToPILImage(),\n",
    "        transforms.Resize((224, 224)),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
    "    ])\n",
    "    image = dlib.load_rgb_image(image_path)\n",
    "    image = trans(image)\n",
    "    image = image.view(1, 3, 224, 224)  # reshape image to match model dimensions (1 batch size)\n",
    "    image = image.to(device)\n",
    "\n",
    "    # fair\n",
    "    with torch.no_grad():\n",
    "        outputs = model_fair_4(image)\n",
    "        outputs = outputs.cpu().detach().numpy()\n",
    "        outputs = np.squeeze(outputs)\n",
    "\n",
    "    gender_outputs = outputs[7:9]\n",
    "    gender_score = np.exp(gender_outputs) / np.sum(np.exp(gender_outputs))\n",
    "    gender_pred = np.argmax(gender_score)\n",
    "    return 1 if gender_pred == 0 else 0"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "separated-limitation",
   "metadata": {},
   "source": [
    "### Results on test set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "id": "behavioral-visit",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 175, FP: 36, TN: 78, FN: 20\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.8187702265372169\n",
      "Precision: 0.8293838862559242\n",
      "Recall: 0.8974358974358975\n",
      "F1_Score: 0.8620689655172414\n"
     ]
    }
   ],
   "source": [
    "acc_fairface, prec_fairface, rec_fairface, f1_fairface = get_scores(get_res_fairface)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "second-burke",
   "metadata": {},
   "source": [
    "### Results on WiderFace"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "solar-chest",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "TP: 459, FP: 135, TN: 365, FN: 36\n",
      "=======================================\n",
      "\u001b[1mMetrics:\u001b[0m\n",
      "Accuracy: 0.828140703517588\n",
      "Precision: 0.7727272727272727\n",
      "Recall: 0.9272727272727272\n",
      "F1_Score: 0.8429752066115703\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.828140703517588, 0.7727272727272727, 0.9272727272727272, 0.8429752066115703)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_scores(get_res_fairface, test_set=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "collected-turning",
   "metadata": {},
   "source": [
    "### Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "id": "appreciated-veteran",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1mTime:\u001b[0m\n",
      "Min time: 0.8669397830963135\n",
      "Max time: 1.6116909980773926\n",
      "Avg time: 0.9408698883056641\n"
     ]
    }
   ],
   "source": [
    "min_time_fairface, max_time_fairface, avg_time_fairface = get_times(get_res_fairface)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "acceptable-cooper",
   "metadata": {},
   "source": [
    "# Results"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "scenic-power",
   "metadata": {},
   "source": [
    "## Results on WiderFace"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "indoor-precipitation",
   "metadata": {},
   "source": [
    "| Method          | Trained on | Precision | Recall  | F1    |\n",
    "|-----------------|------------|-----------|---------|-------|\n",
    "| EfficientNet-B0 | Our        | 0.962     |0.97   |0.966|\n",
    "| EfficientNet-B2 | Our        | 0.978     |0.97   | 0.974 |\n",
    "| **EfficientNet-B4** | **Our**        | **0.974**     | **0.974**  | **0.974**   |\n",
    "| Gil Levi model  | Adience    |0.745     | 0.735    |0.740|\n",
    "| VGG-Face        | IMDb-WIKI  |0.581     | 0.97   | 0.727|\n",
    "| FairFace        | FairFace   |0.773    | 0.927   | 0.843 |"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fixed-broadcasting",
   "metadata": {},
   "source": [
    "## Results on forensic set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "precise-philosophy",
   "metadata": {},
   "source": [
    "| Method          | Trained on | Precision | Recall  | F1    |\n",
    "|-----------------|------------|-----------|---------|-------|\n",
    "| EfficientNet-B0 | Our        | 0.924     | 0.872   | 0.897 |\n",
    "| EfficientNet-B2 | Our        | 0.911     | 0.887   | 0.899 |\n",
    "| **EfficientNet-B4** | **Our**        | **0.902**     | **0.897**  | **0.9**   |\n",
    "| Gil Levi model  | Adience    | 0.893     | 0.6     | 0.718 |\n",
    "| VGG-Face        | IMDb-WIKI  | 0.649     | 0.985   | 0.782 |\n",
    "| FairFace        | FairFace   | 0.829     | 0.897   | 0.862 |"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bacterial-cornell",
   "metadata": {},
   "source": [
    "## Inference latency"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "undefined-syndication",
   "metadata": {},
   "source": [
    "| Method          | Min time | Max time | Avg time |\n",
    "|-----------------|-----------|----------|----------|\n",
    "| EfficientNet-B0 | 0.24     | 0.39     | 0.28     |\n",
    "| EfficientNet-B2 | 0.39     | 1.2      | 0.51     |\n",
    "| EfficientNet-B4 | 0.76     | 1.66     | 0.87     |\n",
    "| **Gil Levi model**  | **0.12**     |**0.24**     | **0.13**     |\n",
    "| VGG-Face        | 1.57     | 2.87     | 1.75     |\n",
    "| FairFace        | 0.87     | 1.61     | 0.94     |"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "underlying-lender",
   "metadata": {},
   "source": [
    "## ROC-AUC on forensic set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 214,
   "id": "guilty-partner",
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "testy_bo, probs_b0 = get_probs(get_res_b0)\n",
    "testy_b2, probs_b2 = get_probs(get_res_b2)\n",
    "testy_b4, probs_b4 = get_probs(get_res_b4)\n",
    "testy_g, probs_g = get_probs(get_res_adience)\n",
    "testy_vgg, probs_vgg = get_probs(get_res_vgg_imdb)\n",
    "testy_f, probs_f = get_probs(get_res_fairface)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "id": "worst-blast",
   "metadata": {},
   "outputs": [],
   "source": [
    "fpr_b0, tpr_b0, auc_b0 = get_fpr_tpr_auc(testy_bo, probs_b0)\n",
    "fpr_b2, tpr_b2, auc_b2 = get_fpr_tpr_auc(testy_b2, probs_b2)\n",
    "fpr_b4, tpr_b4, auc_b4 = get_fpr_tpr_auc(testy_b4, probs_b4)\n",
    "fpr_g, tpr_g, auc_g = get_fpr_tpr_auc(testy_g, probs_g)\n",
    "fpr_vgg, tpr_vgg, auc_vgg = get_fpr_tpr_auc(testy_vgg, probs_vgg)\n",
    "fpr_f, tpr_f, auc_f = get_fpr_tpr_auc(testy_f, probs_f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 216,
   "id": "enhanced-above",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "roc_dict = {\"EfficientNet-B0\" : [fpr_b0, tpr_b0, auc_b0],\n",
    "           \"EfficientNet-B2\" : [fpr_b2, tpr_b2, auc_b2],\n",
    "           \"EfficientNet-B4\" : [fpr_b4, tpr_b4, auc_b4],\n",
    "           \"Gil Levi model\" : [fpr_g, tpr_g, auc_g],\n",
    "           \"VGG-Face\" : [fpr_vgg, tpr_vgg, auc_vgg],\n",
    "           \"FairFace\" : [fpr_f, tpr_f, auc_f]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "id": "successful-absence",
   "metadata": {
    "code_folding": [
     3
    ]
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure()\n",
    "fig.set_size_inches(9, 5)\n",
    "    \n",
    "for method in roc_dict:\n",
    "    fpr, tpr, roc_auc = roc_dict[method]\n",
    "    plt.plot(fpr, tpr,\n",
    "         label=method + ' (AUC = %0.2f)' % roc_auc)\n",
    "\n",
    "plt.plot([0, 1], [0, 1], color='navy', linestyle='--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('ROC')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.grid(True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "encouraging-bracelet",
   "metadata": {},
   "source": [
    "# Helpers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "sunset-screw",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "def convert2RGB(img):\n",
    "    return cv2.cvtColor(img, cv2.COLOR_BGR2RGB)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 182,
   "id": "retained-representative",
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_fpr_tpr_auc(testy, probs):\n",
    "    probs = np.array(probs)[:, 1]\n",
    "    fpr, tpr, treshold = roc_curve(testy, probs)\n",
    "    roc_auc = auc(fpr, tpr)\n",
    "    return fpr, tpr, roc_auc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "id": "pursuant-orleans",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "def get_probs(get_res):\n",
    "    testy = []\n",
    "    probs = []\n",
    "    \n",
    "    # males\n",
    "    for image_name in os.listdir(male_test_dir):\n",
    "        image_path = male_test_dir + image_name\n",
    "        try:\n",
    "            res = get_res(image_path)\n",
    "        except: \n",
    "            print(\"Exception on \" + image_path)\n",
    "            continue\n",
    "        testy.append(1)\n",
    "        if (res >= 0.5):            \n",
    "            probs.append([0, (res - 0.5) * 2])\n",
    "        else: \n",
    "            probs.append([(0.5 - res) * 2, 0])\n",
    "    # females\n",
    "    for image_name in os.listdir(female_test_dir):\n",
    "        image_path = female_test_dir + image_name\n",
    "        try:\n",
    "            res = get_res(image_path)\n",
    "        except: \n",
    "            print(\"Exception on \" + image_path)\n",
    "            continue\n",
    "        testy.append(0)\n",
    "        if (res < 0.5):\n",
    "            probs.append([(0.5 - res) * 2, 0])\n",
    "        else: \n",
    "            probs.append([0, (res - 0.5) * 2])\n",
    "    \n",
    "    return testy, probs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "charming-contribution",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "def get_scores(get_res, test_set=True):\n",
    "    \"\"\"\n",
    "    get_res -- function with input 'image_path' and output -- result of analysis\n",
    "    \"\"\"\n",
    "    TP = 0\n",
    "    FP = 0\n",
    "    TN = 0\n",
    "    FN = 0    \n",
    "\n",
    "    male_dir = male_test_dir if test_set else male_widerfaces_dir\n",
    "    female_dir = male_test_dir if test_set else female_widerfaces_dir\n",
    "    \n",
    "    # males\n",
    "    fst_faces = os.listdir(male_dir)\n",
    "    shuffle(fst_faces)\n",
    "    fst_faces = fst_faces[:500]\n",
    "    for image_name in fst_faces:\n",
    "        image_path = male_dir + image_name\n",
    "        try:\n",
    "            res = get_res(image_path)\n",
    "        except: \n",
    "            continue\n",
    "        if (res >= 0.5):\n",
    "            TP += 1\n",
    "        else: \n",
    "            FN += 1\n",
    "    # females\n",
    "    fst_faces = os.listdir(female_dir)\n",
    "    shuffle(fst_faces)\n",
    "    fst_faces = fst_faces[:500]\n",
    "    for image_name in fst_faces:\n",
    "        image_path = female_dir + image_name\n",
    "        try:\n",
    "            res = get_res(image_path)\n",
    "        except: \n",
    "            continue\n",
    "        if (res < 0.5):\n",
    "            TN += 1\n",
    "        else: \n",
    "            FP += 1\n",
    "            \n",
    "    acc = accuracy(TP, TN, FP, FN)\n",
    "    prec = precision(TP, FP)\n",
    "    rec = recall(TP, FN)\n",
    "    f1 = f_1(prec, rec)\n",
    "    \n",
    "    print(f'TP: {TP}, FP: {FP}, TN: {TN}, FN: {FN}')\n",
    "    print('=======================================')\n",
    "    print(color.BOLD + 'Metrics:' + color.END)\n",
    "    print(f'Accuracy: {acc}')\n",
    "    print(f'Precision: {prec}')\n",
    "    print(f'Recall: {rec}')\n",
    "    print(f'F1_Score: {f1}')\n",
    "    \n",
    "    return acc, prec, rec, f1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "extra-england",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "def get_times(get_res):\n",
    "    times = []\n",
    "    test_image_path = 'TestDataset/CroppedFaces/1/0_m.jpg'\n",
    "    for i in range(0, 1000):\n",
    "        start_time = time.time()\n",
    "        get_res(test_image_path)\n",
    "        ellapsed = time.time() - start_time\n",
    "        times.append(ellapsed)\n",
    "    \n",
    "    min_time = min(times)\n",
    "    max_time = max(times)\n",
    "    avg_time = sum(times)/len(times)\n",
    "    \n",
    "    print(color.BOLD + 'Time:' + color.END)\n",
    "    print(f'Min time: {min_time}')\n",
    "    print(f'Max time: {max_time}')\n",
    "    print(f'Avg time: {avg_time}')\n",
    "    \n",
    "    return min_time, max_time, avg_time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "induced-spice",
   "metadata": {
    "code_folding": [
     0
    ]
   },
   "outputs": [],
   "source": [
    "class color:\n",
    "   PURPLE = '\\033[95m'\n",
    "   CYAN = '\\033[96m'\n",
    "   DARKCYAN = '\\033[36m'\n",
    "   BLUE = '\\033[94m'\n",
    "   GREEN = '\\033[92m'\n",
    "   YELLOW = '\\033[93m'\n",
    "   RED = '\\033[91m'\n",
    "   BOLD = '\\033[1m'\n",
    "   UNDERLINE = '\\033[4m'\n",
    "   END = '\\033[0m'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "answering-pantyhose",
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(TP, TN, FP, FN):\n",
    "    return (TP + TN)/(TP + TN + FP + FN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "nominated-denver",
   "metadata": {},
   "outputs": [],
   "source": [
    "def precision(TP, FP):\n",
    "    return TP/(TP + FP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "congressional-reflection",
   "metadata": {},
   "outputs": [],
   "source": [
    "def recall(TP, FN):\n",
    "    return TP/(TP + FN)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "lined-corpus",
   "metadata": {},
   "outputs": [],
   "source": [
    "def f_1(precision, recall):\n",
    "    return 2 * (precision * recall) / (precision + recall)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 110,
   "id": "artificial-morgan",
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "def build_score_board():\n",
    "    df = pd.DataFrame([['EfficientNet-B0','Our', '0.874', '0.924', '0.872', '0.897'], \n",
    "                      ['EfficientNet-B2', 'Our', '0.874', '0.911', '0.887', '0.899'],\n",
    "                      ['EfficientNet-B4', 'Our', '0.874', '0.902', '0.897', '0.9'],\n",
    "                      ['Gil Levi model', 'Adience', '0.702', '0.893', '0.6', '0.718'],\n",
    "                      ['VGG-Face', 'IMDb-Wiki', '0.654', '0.649', '0.985', '0.782'],\n",
    "                      ['FairFace', 'FairFace', '0.819', '0.829', '0.897', '0.862']],\n",
    "              columns=['Method', 'Trained on', 'Accuracy', 'Precision', 'Recall', 'F1'])    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "id": "deadly-disco",
   "metadata": {
    "code_folding": [
     1
    ]
   },
   "outputs": [],
   "source": [
    "def build_time_board():\n",
    "    df = pd.DataFrame([['EfficientNet-B0', '0.24', '0.39', '0.28'], \n",
    "                      ['EfficientNet-B2', '0.39', '1.2', '0.51'],\n",
    "                      ['EfficientNet-B4', '0.76', '1.66', '0.87'],\n",
    "                      ['Gil Levi model', '0.12', '0.24', '0.13'],\n",
    "                      ['VGG-Face', '1.57', '2.87', '1.75'],\n",
    "                      ['FairFace', '0.87', '1.61', '0.94']],\n",
    "              columns=['Method', 'Min time', 'Max time', 'Avg time'])    \n",
    "    return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "adjustable-russell",
   "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": "3.7.9"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "291.333px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}