SphereFace.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "06aabb63",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:50:10.763028Z",
     "start_time": "2021-09-20T15:50:07.197029Z"
    },
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "import pretrainedmodels\n",
    "\n",
    "from utils.train import train\n",
    "from utils.evaluation import evaluation\n",
    "from utils.loader import get_loader, get_image_list\n",
    "from utils.losses import CostSensitiveRegularizedLoss\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\", category=UserWarning)\n",
    "warnings.filterwarnings(\"ignore\", category=DeprecationWarning)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "cb64b0db",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:50:10.778030Z",
     "start_time": "2021-09-20T15:50:10.764029Z"
    },
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "# hyper params\n",
    "num_of_class = 102\n",
    "learning_rate = 1e-04\n",
    "batch_size = 128\n",
    "input_size = 224\n",
    "device = torch.device(\"cuda\")\n",
    "test_size = 0.2\n",
    "train_size = 0.7\n",
    "val_size = 0.1\n",
    "directoryAgeDB = 'AgeDB/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6d920459",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:50:10.794030Z",
     "start_time": "2021-09-20T15:50:10.779029Z"
    },
    "code_folding": [],
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "def findAge(image_list):\n",
    "    listofzeros = [0] * 102\n",
    "    same_age = [i for i in range(0, 102)]\n",
    "    age_number = dict(zip(same_age, listofzeros))\n",
    "    sum = 0\n",
    "\n",
    "    for i in range(len(image_list)):\n",
    "\n",
    "        if image_list[i]['age'] in age_number:\n",
    "            age_number[image_list[i]['age']] += 1\n",
    "            sum += 1\n",
    "\n",
    "    print(sum)\n",
    "    return dict(age_number)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f8ae2b44",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:51:38.310917Z",
     "start_time": "2021-09-20T15:50:10.795030Z"
    },
    "hide_input": false
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2bcb82c69c9547ccb3e48822cc4955b6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/11541 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Set Length:  34623\n",
      "Test Set Length:  3298\n",
      "Validation Set Length:  1649\n",
      "Total:  39570\n"
     ]
    }
   ],
   "source": [
    "train_image_list, test_image_list, validate_image_list = get_image_list(\n",
    "    train_size=train_size, \n",
    "    test_size=test_size, \n",
    "    val_size=val_size, \n",
    "    directory=directoryAgeDB,\n",
    ")\n",
    "\n",
    "train_loader, test_loader, validation_loader = get_loader(\n",
    "    input_size=input_size,\n",
    "    train_image_list=train_image_list,\n",
    "    test_image_list=test_image_list, \n",
    "    validate_image_list=validate_image_list, \n",
    "    batch_size=batch_size,\n",
    "    train_augment=3,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "5641b0b5",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:51:41.483913Z",
     "start_time": "2021-09-20T15:51:38.312914Z"
    },
    "code_folding": [],
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "resnet1 = pretrainedmodels.resnet18(pretrained='imagenet')\n",
    "resnet1.last_linear = nn.Linear(\n",
    "    in_features=resnet1.last_linear.in_features, \n",
    "    out_features=num_of_class, \n",
    "    bias=False\n",
    ")\n",
    "resnet2 = pretrainedmodels.resnet18(pretrained='imagenet')\n",
    "resnet2.last_linear = nn.Linear(\n",
    "    in_features=resnet2.last_linear.in_features, \n",
    "    out_features=num_of_class, \n",
    "    bias=False\n",
    ")\n",
    "resnet1 = resnet1.to(device)\n",
    "resnet2 = resnet2.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "3ca9b2a2",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:51:41.547913Z",
     "start_time": "2021-09-20T15:51:41.484914Z"
    },
    "hide_input": false
   },
   "outputs": [],
   "source": [
    "from pytorch_metric_learning.losses import SphereFaceLoss\n",
    "criteria = SphereFaceLoss(num_classes=num_of_class, embedding_size=num_of_class)\n",
    "optimizer = torch.optim.Adam(resnet1.parameters(), lr=learning_rate, weight_decay=1e-1)\n",
    "model_name = 'sphereface-aug.pth'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "6ed4bb3e",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:58:56.717803Z",
     "start_time": "2021-09-20T15:51:41.549914Z"
    },
    "hide_input": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 1/5, Train Loss: 5.672906119005266,  Validation Loss: 4.748769173255334\n",
      "Validation Loss Decreased(inf ---> 4.748769173255334)\n",
      "Epoch: 2/5, Train Loss: 4.736126952505639,  Validation Loss: 4.7393861917349005\n",
      "Validation Loss Decreased(4.748769173255334 ---> 4.7393861917349005)\n",
      "Epoch: 3/5, Train Loss: 4.728127349346766,  Validation Loss: 4.705525838411772\n",
      "Validation Loss Decreased(4.7393861917349005 ---> 4.705525838411772)\n",
      "Epoch: 4/5, Train Loss: 4.6991641389487855,  Validation Loss: 4.723685447986309\n",
      "Epoch: 5/5, Train Loss: 4.682548040833421,  Validation Loss: 4.688868559323824\n",
      "Validation Loss Decreased(4.705525838411772 ---> 4.688868559323824)\n"
     ]
    },
    {
     "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": [
    "train(resnet1, optimizer, criteria, train_loader, validation_loader, num_of_epoch=5, model_name=model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "10b735ef",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:59:37.834802Z",
     "start_time": "2021-09-20T15:58:56.719801Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 19.155548095703125\n",
      "Minimum: 0.0, Maximum: 73.0, Median: 16.0\n",
      "Mean Absolute Error: 19.04196548461914\n",
      "Minimum: 0.0, Maximum: 68.0, Median: 16.0\n"
     ]
    }
   ],
   "source": [
    "resnet1.load_state_dict(torch.load(model_name))\n",
    "loss_per_class = evaluation(resnet1, test_loader)\n",
    "_ = evaluation(resnet1, train_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "6452b440",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T15:59:37.895799Z",
     "start_time": "2021-09-20T15:59:37.835801Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3298\n",
      "Age: Test Sample: Loss\n",
      "1 : 1 : 26.00\n",
      "2  No Sample\n",
      "3 : 2 : 24.00\n",
      "4 : 1 : 23.00\n",
      "5 : 1 : 22.00\n",
      "6 : 2 : 30.00\n",
      "7 : 2 : 20.00\n",
      "8 : 2 : 19.00\n",
      "9 : 1 : 18.00\n",
      "10 : 3 : 17.00\n",
      "11 : 4 : 20.50\n",
      "12 : 3 : 15.00\n",
      "13 : 4 : 14.00\n",
      "14 : 5 : 13.00\n",
      "15 : 9 : 12.00\n",
      "16 : 10 : 15.10\n",
      "17 : 16 : 15.12\n",
      "18 : 24 : 12.75\n",
      "19 : 22 : 10.68\n",
      "20 : 32 : 12.22\n",
      "21 : 33 : 7.64\n",
      "22 : 37 : 6.46\n",
      "23 : 42 : 7.86\n",
      "24 : 52 : 5.87\n",
      "25 : 57 : 5.33\n",
      "26 : 55 : 4.45\n",
      "27 : 61 : 3.20\n",
      "28 : 69 : 3.78\n",
      "29 : 62 : 4.19\n",
      "30 : 77 : 5.56\n",
      "31 : 63 : 6.38\n",
      "32 : 75 : 7.05\n",
      "33 : 67 : 6.99\n",
      "34 : 72 : 7.68\n",
      "35 : 82 : 8.49\n",
      "36 : 69 : 9.00\n",
      "37 : 74 : 9.65\n",
      "38 : 76 : 10.26\n",
      "39 : 64 : 10.69\n",
      "40 : 77 : 11.68\n",
      "41 : 64 : 12.98\n",
      "42 : 71 : 13.28\n",
      "43 : 65 : 13.49\n",
      "44 : 56 : 14.61\n",
      "45 : 68 : 14.99\n",
      "46 : 57 : 16.84\n",
      "47 : 58 : 17.83\n",
      "48 : 62 : 17.52\n",
      "49 : 53 : 18.21\n",
      "50 : 63 : 19.13\n",
      "51 : 57 : 17.28\n",
      "52 : 62 : 21.77\n",
      "53 : 50 : 22.36\n",
      "54 : 51 : 21.96\n",
      "55 : 59 : 22.56\n",
      "56 : 45 : 25.29\n",
      "57 : 49 : 25.76\n",
      "58 : 49 : 26.96\n",
      "59 : 43 : 30.33\n",
      "60 : 49 : 29.86\n",
      "61 : 41 : 29.80\n",
      "62 : 48 : 30.83\n",
      "63 : 44 : 30.16\n",
      "64 : 39 : 33.51\n",
      "65 : 46 : 32.91\n",
      "66 : 35 : 32.63\n",
      "67 : 35 : 37.43\n",
      "68 : 43 : 38.91\n",
      "69 : 30 : 36.87\n",
      "70 : 43 : 38.70\n",
      "71 : 29 : 38.69\n",
      "72 : 35 : 38.69\n",
      "73 : 31 : 40.45\n",
      "74 : 28 : 39.39\n",
      "75 : 32 : 43.34\n",
      "76 : 22 : 41.18\n",
      "77 : 21 : 45.48\n",
      "78 : 21 : 45.38\n",
      "79 : 15 : 48.40\n",
      "80 : 18 : 49.72\n",
      "81 : 15 : 44.40\n",
      "82 : 14 : 49.86\n",
      "83 : 11 : 52.27\n",
      "84 : 9 : 52.44\n",
      "85 : 12 : 53.50\n",
      "86 : 7 : 50.57\n",
      "87 : 6 : 54.00\n",
      "88 : 4 : 56.50\n",
      "89 : 4 : 53.00\n",
      "90 : 4 : 54.00\n",
      "91 : 3 : 64.00\n",
      "92 : 2 : 65.00\n",
      "93 : 3 : 66.00\n",
      "94 : 1 : 67.00\n",
      "95 : 2 : 56.50\n",
      "96 : 1 : 46.00\n",
      "97 : 1 : 70.00\n",
      "98 : 1 : 71.00\n",
      "99 : 1 : 72.00\n",
      "100 : 1 : 73.00\n",
      "101 : 1 : 56.00\n"
     ]
    }
   ],
   "source": [
    "sample_per_class = findAge(test_image_list)\n",
    "print('Age: Test Sample: Loss')\n",
    "for i in range(1, 102):\n",
    "    if sample_per_class[i] == 0:\n",
    "        print(i,' No Sample')\n",
    "    else:\n",
    "        print(i,':', sample_per_class[i],':', format(loss_per_class[i].item()/sample_per_class[i],'.2f'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "38bb6666",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T16:04:22.057710Z",
     "start_time": "2021-09-20T16:01:40.899534Z"
    }
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a42b9d802dba4b06b6574a65e0c2411b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/11541 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Set Length:  11541\n",
      "Test Set Length:  3298\n",
      "Validation Set Length:  1649\n",
      "Total:  16488\n",
      "Epoch: 1/5, Train Loss: 7.428118946788075,  Validation Loss: 4.907127600449782\n",
      "Validation Loss Decreased(inf ---> 4.907127600449782)\n",
      "Epoch: 2/5, Train Loss: 4.816846606495616,  Validation Loss: 4.794305031116192\n",
      "Validation Loss Decreased(4.907127600449782 ---> 4.794305031116192)\n",
      "Epoch: 3/5, Train Loss: 4.753860861390502,  Validation Loss: 4.7399570391728325\n",
      "Validation Loss Decreased(4.794305031116192 ---> 4.7399570391728325)\n",
      "Epoch: 4/5, Train Loss: 4.7397076690589985,  Validation Loss: 4.802999276381272\n",
      "Epoch: 5/5, Train Loss: 4.740888637500805,  Validation Loss: 4.7230054048391485\n",
      "Validation Loss Decreased(4.7399570391728325 ---> 4.7230054048391485)\n"
     ]
    },
    {
     "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": [
    "train_loader, test_loader, validation_loader = get_loader(\n",
    "    input_size=input_size,\n",
    "    train_image_list=train_image_list,\n",
    "    test_image_list=test_image_list, \n",
    "    validate_image_list=validate_image_list, \n",
    "    batch_size=batch_size,\n",
    "    train_augment=1,\n",
    ")\n",
    "criteria = SphereFaceLoss(num_classes=num_of_class, embedding_size=num_of_class)\n",
    "optimizer = torch.optim.Adam(resnet2.parameters(), lr=learning_rate, weight_decay=1e-1)\n",
    "model_name = 'sphereface.pth'\n",
    "train(resnet2, optimizer, criteria, train_loader, validation_loader, num_of_epoch=5, model_name=model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "647a89b6",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T16:04:56.402077Z",
     "start_time": "2021-09-20T16:04:26.203451Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean Absolute Error: 31.811098098754883\n",
      "Minimum: 0.0, Maximum: 97.0, Median: 30.0\n",
      "Mean Absolute Error: 31.848365783691406\n",
      "Minimum: 0.0, Maximum: 98.0, Median: 31.0\n"
     ]
    }
   ],
   "source": [
    "resnet2.load_state_dict(torch.load(model_name))\n",
    "loss_per_class = evaluation(resnet2, test_loader)\n",
    "_ = evaluation(resnet2, train_loader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "4cac931a",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2021-09-20T16:05:10.894722Z",
     "start_time": "2021-09-20T16:05:10.797723Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3298\n",
      "Age: Test Sample: Loss\n",
      "1 : 1 : 59.00\n",
      "2  No Sample\n",
      "3 : 2 : 38.50\n",
      "4 : 1 : 97.00\n",
      "5 : 1 : 18.00\n",
      "6 : 2 : 76.50\n",
      "7 : 2 : 73.00\n",
      "8 : 2 : 20.50\n",
      "9 : 1 : 64.00\n",
      "10 : 3 : 75.33\n",
      "11 : 4 : 47.25\n",
      "12 : 3 : 60.00\n",
      "13 : 4 : 48.75\n",
      "14 : 5 : 27.60\n",
      "15 : 9 : 43.67\n",
      "16 : 10 : 38.30\n",
      "17 : 16 : 53.06\n",
      "18 : 24 : 45.71\n",
      "19 : 22 : 51.27\n",
      "20 : 32 : 51.47\n",
      "21 : 33 : 37.36\n",
      "22 : 37 : 38.65\n",
      "23 : 42 : 48.12\n",
      "24 : 52 : 41.23\n",
      "25 : 57 : 36.07\n",
      "26 : 55 : 43.96\n",
      "27 : 61 : 34.67\n",
      "28 : 69 : 38.93\n",
      "29 : 62 : 40.66\n",
      "30 : 77 : 36.45\n",
      "31 : 63 : 40.02\n",
      "32 : 75 : 39.23\n",
      "33 : 67 : 34.96\n",
      "34 : 72 : 37.75\n",
      "35 : 82 : 31.06\n",
      "36 : 69 : 33.78\n",
      "37 : 74 : 31.66\n",
      "38 : 76 : 32.97\n",
      "39 : 64 : 32.77\n",
      "40 : 77 : 33.17\n",
      "41 : 64 : 31.16\n",
      "42 : 71 : 31.49\n",
      "43 : 65 : 30.22\n",
      "44 : 56 : 33.52\n",
      "45 : 68 : 29.97\n",
      "46 : 57 : 33.23\n",
      "47 : 58 : 27.86\n",
      "48 : 62 : 28.52\n",
      "49 : 53 : 29.28\n",
      "50 : 63 : 31.70\n",
      "51 : 57 : 30.95\n",
      "52 : 62 : 29.66\n",
      "53 : 50 : 29.22\n",
      "54 : 51 : 29.04\n",
      "55 : 59 : 27.36\n",
      "56 : 45 : 25.16\n",
      "57 : 49 : 25.51\n",
      "58 : 49 : 26.16\n",
      "59 : 43 : 24.63\n",
      "60 : 49 : 29.39\n",
      "61 : 41 : 23.85\n",
      "62 : 48 : 23.27\n",
      "63 : 44 : 19.48\n",
      "64 : 39 : 25.00\n",
      "65 : 46 : 23.22\n",
      "66 : 35 : 17.94\n",
      "67 : 35 : 24.46\n",
      "68 : 43 : 19.26\n",
      "69 : 30 : 16.83\n",
      "70 : 43 : 21.02\n",
      "71 : 29 : 27.86\n",
      "72 : 35 : 21.91\n",
      "73 : 31 : 32.65\n",
      "74 : 28 : 28.43\n",
      "75 : 32 : 20.22\n",
      "76 : 22 : 26.23\n",
      "77 : 21 : 34.62\n",
      "78 : 21 : 31.57\n",
      "79 : 15 : 22.40\n",
      "80 : 18 : 35.39\n",
      "81 : 15 : 15.40\n",
      "82 : 14 : 16.00\n",
      "83 : 11 : 28.00\n",
      "84 : 9 : 26.78\n",
      "85 : 12 : 19.83\n",
      "86 : 7 : 35.86\n",
      "87 : 6 : 39.00\n",
      "88 : 4 : 37.50\n",
      "89 : 4 : 42.50\n",
      "90 : 4 : 25.75\n",
      "91 : 3 : 11.00\n",
      "92 : 2 : 3.00\n",
      "93 : 3 : 21.33\n",
      "94 : 1 : 21.00\n",
      "95 : 2 : 17.50\n",
      "96 : 1 : 4.00\n",
      "97 : 1 : 96.00\n",
      "98 : 1 : 42.00\n",
      "99 : 1 : 30.00\n",
      "100 : 1 : 44.00\n",
      "101 : 1 : 32.00\n"
     ]
    }
   ],
   "source": [
    "sample_per_class = findAge(test_image_list)\n",
    "print('Age: Test Sample: Loss')\n",
    "for i in range(1, 102):\n",
    "    if sample_per_class[i] == 0:\n",
    "        print(i,' No Sample')\n",
    "    else:\n",
    "        print(i,':', sample_per_class[i],':', format(loss_per_class[i].item()/sample_per_class[i],'.2f'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "92c44ac5",
   "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": "3.8.8"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}