get_landmarks_geom.features.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "4ce2fce5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/MEBeauty-database/images/male/indian/puvvukonvict-photography-AtpSEe3yoIg-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/mideastern/imad-clicks-2_qmEnz7bQ4-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/indian/shifaaz-shamoon-MqLy-G-dBi8-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/black/payton-tuttle-n_RdRxH_7h4-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/hispanic/aryo-lahap-ZVc4Zg4TY8s-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/asian/tanjir-ahmed-chowdhury-YMJvX8k3Kfc-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/caucasian/brunette-2478405_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/business-man-1385050_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/boy-1105361_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/hispanic/mateus-campos-felipe-64sktAOWasI-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/caucasian/creating-a-brand-WxAOdbDpMiU-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/asian/pexels-rodnae-productions-4911184.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/indian/pexels-nh-noyon-4250862.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/indian/brown-4197806_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/asian/pexels-kevin-kabore-4966995.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/man-1488305_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/black/doil-oh-v15GcVql4CM-unsplash.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/indian/guy10.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/g_9.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/g_7.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/model-4493512_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/montgomery-clift-402456_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/mideastern/guy14.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/caucasian/girl2.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/black/woman-984426_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/caucasian/esther-williams-402543_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/mideastern/jonathan-borba-5rQG1mib90I-unsplash (1).jpg\n",
      "/home/ubuntu/MEBeauty-database/images/male/caucasian/man-5880780_1920.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/caucasian/charlize-theron-669608_1280.jpg\n",
      "/home/ubuntu/MEBeauty-database/images/female/hispanic/girl1.jpg\n"
     ]
    }
   ],
   "source": [
    "#coding=utf8\n",
    "import cv2  \n",
    "import dlib  \n",
    "import numpy  \n",
    "import sys   \n",
    "df =data    \n",
    "PREDICTOR_PATH =\"/home/ubuntu/shape_predictor_68_face_landmarks.dat\"  \n",
    "  \n",
    "#1.使用dlib自带的frontal_face_detector作为我们的人脸提取器  \n",
    "detector = dlib.get_frontal_face_detector()  \n",
    " \n",
    "#2.使用官方提供的模型构建特征提取器  \n",
    "predictor = dlib.shape_predictor(PREDICTOR_PATH)  \n",
    "  \n",
    "  \n",
    "df2 = pd.DataFrame(columns = ['image', 'score', 'landmarks'])\n",
    "    \n",
    "for index, row in df.iterrows():\n",
    "   \n",
    " try:\n",
    "   im = cv2.imread(row['image'])  \n",
    "  # im = cv2.cvtColor(im, cv2.COLOR_BGR2RGB)\n",
    "   gray = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY)\n",
    "   dets = detector(gray, 0)\n",
    "#3.使用detector进行人脸检测 rects为返回的结果  \n",
    "   #rects = detector(im,1)  \n",
    "   #im = dlib.load_rgb_image(row['image'])\n",
    "#4.输出人脸数，dets的元素个数即为脸的个数  \n",
    "   if len(dets) >= 1:  \n",
    "      #print(\"{} faces detected\".format(len(dets)))  \n",
    "      \n",
    "      \n",
    "    #5.使用predictor进行人脸关键点识别  \n",
    "     # for i in range(0, 68):\n",
    "      #   coords[i] = (shape.part(i).x, shape.part(i).y) \n",
    "      for det in dets:\n",
    "        f=''    \n",
    "        #landmarks = shape_to_np(predictor(gray, det))\n",
    "        landmarks = numpy.matrix([[p.x,p.y] for p in predictor(gray,det).parts()]) \n",
    "\n",
    "  \n",
    "    #使用enumerate 函数遍历序列中的元素以及它们的下标  \n",
    "        for idx,point in enumerate(landmarks):  \n",
    "           pos = (point[0,0],point[0,1])  \n",
    "        \n",
    "           f = f + str(point[0,0]) + ',' + str(point[0,1]) + ','\n",
    "        \n",
    "      #print(landmarks)\n",
    "    #使用enumerate 函数遍历序列中的元素以及它们的下标  \n",
    "        df2.loc[len(df2)]=[row['image'],row['score'], f]\n",
    "    \n",
    " except:\n",
    "        print(row['image'])    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "71189ac2",
   "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>image</th>\n",
       "      <th>score</th>\n",
       "      <th>landmarks</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/male/ind...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>136,291,144,315,153,339,167,364,185,387,203,40...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/male/asi...</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>117,197,116,237,118,281,124,322,140,358,165,39...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/c...</td>\n",
       "      <td>1.006920</td>\n",
       "      <td>139,241,128,264,120,290,118,317,119,345,120,37...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/i...</td>\n",
       "      <td>1.500000</td>\n",
       "      <td>88,185,86,223,89,261,96,300,112,336,133,370,16...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/male/cau...</td>\n",
       "      <td>1.625000</td>\n",
       "      <td>69,180,67,232,68,281,77,331,95,379,119,425,151...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2454</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/b...</td>\n",
       "      <td>9.281250</td>\n",
       "      <td>113,229,118,274,131,320,152,363,177,405,208,44...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2455</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/h...</td>\n",
       "      <td>9.358025</td>\n",
       "      <td>158,234,159,256,161,278,166,299,173,320,186,33...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2456</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/a...</td>\n",
       "      <td>9.375000</td>\n",
       "      <td>238,234,234,250,231,266,233,283,237,299,242,31...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2457</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/i...</td>\n",
       "      <td>9.375000</td>\n",
       "      <td>175,227,172,248,173,270,177,294,185,317,195,33...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2458</th>\n",
       "      <td>/home/ubuntu/MEBeauty-database/images/female/c...</td>\n",
       "      <td>9.625000</td>\n",
       "      <td>169,237,173,259,179,280,193,300,210,319,228,33...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2459 rows × 3 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                  image     score  \\\n",
       "0     /home/ubuntu/MEBeauty-database/images/male/ind...  1.000000   \n",
       "1     /home/ubuntu/MEBeauty-database/images/male/asi...  1.000000   \n",
       "2     /home/ubuntu/MEBeauty-database/images/female/c...  1.006920   \n",
       "3     /home/ubuntu/MEBeauty-database/images/female/i...  1.500000   \n",
       "4     /home/ubuntu/MEBeauty-database/images/male/cau...  1.625000   \n",
       "...                                                 ...       ...   \n",
       "2454  /home/ubuntu/MEBeauty-database/images/female/b...  9.281250   \n",
       "2455  /home/ubuntu/MEBeauty-database/images/female/h...  9.358025   \n",
       "2456  /home/ubuntu/MEBeauty-database/images/female/a...  9.375000   \n",
       "2457  /home/ubuntu/MEBeauty-database/images/female/i...  9.375000   \n",
       "2458  /home/ubuntu/MEBeauty-database/images/female/c...  9.625000   \n",
       "\n",
       "                                              landmarks  \n",
       "0     136,291,144,315,153,339,167,364,185,387,203,40...  \n",
       "1     117,197,116,237,118,281,124,322,140,358,165,39...  \n",
       "2     139,241,128,264,120,290,118,317,119,345,120,37...  \n",
       "3     88,185,86,223,89,261,96,300,112,336,133,370,16...  \n",
       "4     69,180,67,232,68,281,77,331,95,379,119,425,151...  \n",
       "...                                                 ...  \n",
       "2454  113,229,118,274,131,320,152,363,177,405,208,44...  \n",
       "2455  158,234,159,256,161,278,166,299,173,320,186,33...  \n",
       "2456  238,234,234,250,231,266,233,283,237,299,242,31...  \n",
       "2457  175,227,172,248,173,270,177,294,185,317,195,33...  \n",
       "2458  169,237,173,259,179,280,193,300,210,319,228,33...  \n",
       "\n",
       "[2459 rows x 3 columns]"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "8741c963",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/anaconda3/envs/aws_neuron_pytorch_p36/lib/python3.6/site-packages/ipykernel/__main__.py:81: DeprecationWarning: string or file could not be read to its end due to unmatched data; this will raise a ValueError in the future.\n"
     ]
    },
    {
     "ename": "IndexError",
     "evalue": "index 34 is out of bounds for axis 0 with size 0",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mIndexError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-3-bab6aae1b667>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m     82\u001b[0m    \u001b[0mlandmarks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnumpy\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexpand_dims\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlandmarks\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     83\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 84\u001b[0;31m    \u001b[0mfeaturesALL\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgenerateAllFeatures\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlandmarks\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     85\u001b[0m    \u001b[0mdf3\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mloc\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdf3\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mrow\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'image'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mrow\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'score'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mfeaturesALL\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     86\u001b[0m    \u001b[0;31m#print \"Generate Feature Successfully!\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-bab6aae1b667>\u001b[0m in \u001b[0;36mgenerateAllFeatures\u001b[0;34m(allLandmarkCoordinates)\u001b[0m\n\u001b[1;32m     71\u001b[0m                 \u001b[0mi\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mi\u001b[0m\u001b[0;34m+\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     72\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 73\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mgenerateFeatures\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpointIndices1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpointIndices2\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpointIndices3\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpointIndices4\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mallLandmarkCoordinates\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     74\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     75\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m<ipython-input-3-bab6aae1b667>\u001b[0m in \u001b[0;36mgenerateFeatures\u001b[0;34m(pointIndices1, pointIndices2, pointIndices3, pointIndices4, allLandmarkCoordinates)\u001b[0m\n\u001b[1;32m     29\u001b[0m                 \u001b[0mratios\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m;\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     30\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0mi\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpointIndices1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 31\u001b[0;31m                         \u001b[0mx1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlandmarkCoordinates\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpointIndices1\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     32\u001b[0m                         \u001b[0my1\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlandmarkCoordinates\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mpointIndices1\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     33\u001b[0m                         \u001b[0mx2\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mlandmarkCoordinates\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpointIndices2\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mIndexError\u001b[0m: index 34 is out of bounds for axis 0 with size 0"
     ]
    }
   ],
   "source": [
    "import math\n",
    "import numpy\n",
    "import itertools\n",
    "\n",
    "def facialRatio(points):\n",
    "\tx1 = points[0];\n",
    "\ty1 = points[1];\n",
    "\tx2 = points[2];\n",
    "\ty2 = points[3];\n",
    "\tx3 = points[4];\n",
    "\ty3 = points[5];\n",
    "\tx4 = points[6];\n",
    "\ty4 = points[7];\n",
    "\n",
    "\tdist1 = math.sqrt((x1-x2)**2 + (y1-y2)**2)\n",
    "\tdist2 = math.sqrt((x3-x4)**2 + (y3-y4)**2)\n",
    "\n",
    "\tratio = dist1/dist2\n",
    "\n",
    "\treturn ratio\n",
    "\n",
    "\n",
    "def generateFeatures(pointIndices1, pointIndices2, pointIndices3, pointIndices4, allLandmarkCoordinates):\n",
    "\tsize = allLandmarkCoordinates.shape\n",
    "\tallFeatures = numpy.zeros((size[0], len(pointIndices1)))\n",
    "\t#print(size[0])\n",
    "\tfor x in range(0, size[0]):\n",
    "\t\tlandmarkCoordinates = allLandmarkCoordinates[x, :]\n",
    "\t\tratios = [];\n",
    "\t\tfor i in range(0, len(pointIndices1)):\n",
    "\t\t\tx1 = landmarkCoordinates[2*(pointIndices1[i]-1)]\n",
    "\t\t\ty1 = landmarkCoordinates[2*pointIndices1[i] - 1]\n",
    "\t\t\tx2 = landmarkCoordinates[2*(pointIndices2[i]-1)]\n",
    "\t\t\ty2 = landmarkCoordinates[2*pointIndices2[i] - 1]\n",
    "\n",
    "\t\t\tx3 = landmarkCoordinates[2*(pointIndices3[i]-1)]\n",
    "\t\t\ty3 = landmarkCoordinates[2*pointIndices3[i] - 1]\n",
    "\t\t\tx4 = landmarkCoordinates[2*(pointIndices4[i]-1)]\n",
    "\t\t\ty4 = landmarkCoordinates[2*pointIndices4[i] - 1]\n",
    "\n",
    "\t\t\tpoints = [x1, y1, x2, y2, x3, y3, x4, y4]\n",
    "\t\t\tratios.append(facialRatio(points))\n",
    "\t\tallFeatures[x, :] = numpy.asarray(ratios)\n",
    "\treturn allFeatures\n",
    "\n",
    "\n",
    "def generateAllFeatures(allLandmarkCoordinates):\n",
    "\ta = [18, 22, 23, 27, 37, 40, 43, 46, 28, 32, 34, 36, 5, 9, 13, 49, 55, 52, 58]\n",
    "\tcombinations = itertools.combinations(a, 4)\n",
    "\ti = 0\n",
    "\tpointIndices1 = [];\n",
    "\tpointIndices2 = [];\n",
    "\tpointIndices3 = [];\n",
    "\tpointIndices4 = [];\n",
    "\n",
    "\tfor combination in combinations:\n",
    "\t\tpointIndices1.append(combination[0])\n",
    "\t\tpointIndices2.append(combination[1])\n",
    "\t\tpointIndices3.append(combination[2])\n",
    "\t\tpointIndices4.append(combination[3])\n",
    "\t\ti = i+1\n",
    "\t\tpointIndices1.append(combination[0])\n",
    "\t\tpointIndices2.append(combination[2])\n",
    "\t\tpointIndices3.append(combination[1])\n",
    "\t\tpointIndices4.append(combination[3])\n",
    "\t\ti = i+1\n",
    "\t\tpointIndices1.append(combination[0])\n",
    "\t\tpointIndices2.append(combination[3])\n",
    "\t\tpointIndices3.append(combination[1])\n",
    "\t\tpointIndices4.append(combination[2])\n",
    "\t\ti = i+1\n",
    "\n",
    "\treturn generateFeatures(pointIndices1, pointIndices2, pointIndices3, pointIndices4, allLandmarkCoordinates)\n",
    "\n",
    "\n",
    "df3 = pd.DataFrame(columns = ['image', 'score', 'features'])\n",
    "    \n",
    "for index, row in df.iterrows(): \n",
    "\n",
    "   #features = numpy.loadtxt(row['landmarks'], dtype=int, usecols=range(136))\n",
    "   landmarks = numpy.fromstring(row['landmarks'], dtype=int, sep=',')\n",
    "   landmarks = numpy.expand_dims(landmarks, axis=0) \n",
    "    \n",
    "   featuresALL = generateAllFeatures(landmarks)\n",
    "   df3.loc[len(df3)]=[row['image'],row['score'], featuresALL]\n",
    "   #print \"Generate Feature Successfully!\"\n",
    "#pointIndices1 = [20, 20, 45, 45]\n",
    "#pointIndices2 = [58, 9, 58, 58]\n",
    "#pointIndices3 = [5, 7, 5, 32]\n",
    "#pointIndices4 = [13, 13, 11, 36]\n",
    "#features = generateFeatures(pointIndices1, pointIndices2, pointIndices3, pointIndices4, landmarks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "a1959c4a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "df = pd.read_csv('landmarks.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "e7a712e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from scipy import spatial\n",
    "import os\n",
    "#prepare data with generic scores\n",
    "\n",
    "os. chdir('/home/ubuntu/MEBeauty-database/')\n",
    "\n",
    "txt1 = pd.read_csv('scores/train_2022.txt', sep=\" \",  header = None)\n",
    "txt2 = pd.read_csv('scores/test_2022.txt', sep=\" \",  header = None)\n",
    "txt3 = pd.read_csv('scores/val_2022.txt', sep=\" \", header = None)\n",
    "data = pd.concat([txt1,txt2,txt3])\n",
    "data.columns = ['image', 'score']\n",
    "data.image  = data.image.str.replace('/home/ubuntu/ME-beautydatabase/images/','/home/ubuntu/MEBeauty-database/images/')\n",
    "\n",
    "\n",
    "data = data.sort_values(by=['score'])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Environment (conda_aws_neuron_pytorch_p36)",
   "language": "python",
   "name": "conda_aws_neuron_pytorch_p36"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}