Fixing hopenet

natanielruiz

2017-09-14 dd62d6fa4a85f18a29de009a972f5599b19ec946

Fixing hopenet

 code/batch_testing.py

@@ -123,9 +123,9 @@
            label_roll = labels[:,2].float()

            pre_yaw, pre_pitch, pre_roll, angles = model(images)
            yaw = angles[args.iter_ref-1][:,0].cpu().data
            pitch = angles[args.iter_ref-1][:,1].cpu().data
            roll = angles[args.iter_ref-1][:,2].cpu().data
            yaw = angles[args.iter_ref][:,0].cpu().data
            pitch = angles[args.iter_ref][:,1].cpu().data
            roll = angles[args.iter_ref][:,2].cpu().data

            # Mean absolute error
            yaw_error += torch.sum(torch.abs(yaw - label_yaw) * 3)

 code/test.py

@@ -110,11 +110,12 @@
        label_roll = labels[:,2].float()

        pre_yaw, pre_pitch, pre_roll, angles = model(images)
        yaw = angles[args.iter_ref-1][:,0].cpu().data
        pitch = angles[args.iter_ref-1][:,1].cpu().data
        roll = angles[args.iter_ref-1][:,2].cpu().data
        yaw = angles[args.iter_ref][:,0].cpu().data
        pitch = angles[args.iter_ref][:,1].cpu().data
        roll = angles[args.iter_ref][:,2].cpu().data

        # Mean absolute error
        print yaw.numpy(), label_yaw.numpy()
        yaw_error += torch.sum(torch.abs(yaw - label_yaw) * 3)
        pitch_error += torch.sum(torch.abs(pitch - label_pitch) * 3)
        roll_error += torch.sum(torch.abs(roll - label_roll) * 3)

 code/train.py

@@ -241,7 +241,7 @@

            # Finetuning loss
            loss_seq = [loss_yaw, loss_pitch, loss_roll]
            for idx in xrange(args.iter_ref):
            for idx in xrange(args.iter_ref+1):
                loss_angles = reg_criterion(angles[idx], label_angles.float())
                loss_seq.append(loss_angles)


 practice/.ipynb_checkpoints/load_AFLW-Copy1-checkpoint.ipynb

@@ -2,7 +2,7 @@
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
@@ -23,7 +23,7 @@
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
@@ -36,20 +36,16 @@
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-1f2606c2a679>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     34\u001b[0m     \u001b[0;32mif\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0misfile\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_path\u001b[0m\u001b[0;34m)\u001b[0m  \u001b[0;34m==\u001b[0m \u001b[0mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     35\u001b[0m         \u001b[0;31m#image = cv2.imread(input_path, 0) #load in grayscale\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 36\u001b[0;31m         \u001b[0mimage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcv2\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimread\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     37\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     38\u001b[0m         \u001b[0;31m#Image dimensions\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done\n"
     ]
    }
   ],
@@ -148,11 +144,19 @@
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test\n"
     ]
    }
   ],
   "source": [
    "print 'test'"
   ]

 practice/.ipynb_checkpoints/load_AFLW-KEPLER_split-checkpoint.ipynb

New file
@@ -0,0 +1,225 @@
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import torch\n",
    "from torch.utils.serialization import load_lua\n",
    "import os\n",
    "import scipy.io as sio\n",
    "import cv2\n",
    "import math\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "import sqlite3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Change this paths according to your directories\n",
    "images_path = \"/Data/nruiz9/data/facial_landmarks/AFLW/aflw/data/flickr/\"\n",
    "storing_path = \"/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(573, 1)\n",
      "(232, 1)\n",
      "(165, 1)\n",
      "3\n"
     ]
    }
   ],
   "source": [
    "#Load KEPLER split file\n",
    "test_set = sio.loadmat('/Data/nruiz9/data/facial_landmarks/AFLW/testset.mat')\n",
    "print test_set['test'][0][0][0].shape\n",
    "print test_set['test'][0][0][1].shape\n",
    "print test_set['test'][0][0][2].shape\n",
    "print len(test_set['test'][0][0])\n",
    "\n",
    "# I will just use every single one for now. If results are not good I'll sample equally."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done\n"
     ]
    }
   ],
   "source": [
    "#Image counter\n",
    "counter = 1\n",
    "\n",
    "#Open the sqlite database\n",
    "conn = sqlite3.connect('/Data/nruiz9/data/facial_landmarks/AFLW/aflw/data/aflw.sqlite')\n",
    "c = conn.cursor()\n",
    "\n",
    "#Creating the query string for retriving: roll, pitch, yaw and faces position\n",
    "#Change it according to what you want to retrieve\n",
    "select_string = \"faceimages.filepath, faces.face_id, facepose.roll, facepose.pitch, facepose.yaw, facerect.x, facerect.y, facerect.w, facerect.h\"\n",
    "from_string = \"faceimages, faces, facepose, facerect\"\n",
    "where_string = \"faces.face_id = facepose.face_id and faces.file_id = faceimages.file_id and faces.face_id = facerect.face_id\"\n",
    "query_string = \"SELECT \" + select_string + \" FROM \" + from_string + \" WHERE \" + where_string\n",
    "\n",
    "#It iterates through the rows returned from the query\n",
    "for row in c.execute(query_string):\n",
    "\n",
    "    #Using our specific query_string, the \"row\" variable will contain:\n",
    "    # row[0] = image path\n",
    "    # row[1] = face id\n",
    "    # row[2] = rollgma\n",
    "    # row[3] = pitch\n",
    "    # row[4] = yaw\n",
    "    # row[5] = face coord x\n",
    "    # row[6] = face coord y\n",
    "    # row[7] = face width\n",
    "    # row[8] = face heigh\n",
    "\n",
    "    #Creating the full path names for input and output\n",
    "    input_path = images_path + str(row[0])\n",
    "    output_path = storing_path + str(row[0])\n",
    "\n",
    "    #If the file exist then open it       \n",
    "    if(os.path.isfile(input_path)  == True):\n",
    "        #image = cv2.imread(input_path, 0) #load in grayscale\n",
    "        image = cv2.imread(input_path)\n",
    "\n",
    "        #Image dimensions\n",
    "        image_h, image_w, _ = image.shape\n",
    "        #Roll, pitch and yaw\n",
    "        roll   = row[2]\n",
    "        pitch  = row[3]\n",
    "        yaw    = row[4]\n",
    "        #Face rectangle coords\n",
    "        face_x = row[5]\n",
    "        face_y = row[6]\n",
    "        face_w = row[7]\n",
    "        face_h = row[8]\n",
    "\n",
    "        #Error correction\n",
    "        k = 0.35\n",
    "        x_min = face_x - face_w * k * 0.6\n",
    "        x_max = face_x + face_w + face_w * k * 0.6\n",
    "        y_min = face_y - face_h * k * 2\n",
    "        y_max = face_y + face_h + face_h * k * 0.6\n",
    "        \n",
    "        x_min = int(max(0, x_min))\n",
    "        x_max = int(min(image_w, x_max))\n",
    "        y_min = int(max(0, y_min))\n",
    "        y_max = int(min(image_h, y_max))\n",
    "        \n",
    "        #Crop the face from the image\n",
    "        image_cropped = np.copy(image[y_min:y_max, x_min:x_max])\n",
    "        #Uncomment the lines below if you want to rescale the image to a particular size\n",
    "        to_size = 240\n",
    "        image_cropped = cv2.resize(image_cropped, (to_size,to_size), interpolation = cv2.INTER_AREA)\n",
    "        #Uncomment the line below if you want to use adaptive histogram normalisation\n",
    "        #clahe = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(5,5))\n",
    "        #image_normalised = clahe.apply(image_rescaled)\n",
    "        #Save the image\n",
    "        #change \"image_cropped\" with the last uncommented variable name above\n",
    "        cv2.imwrite(output_path, image_cropped)\n",
    "        \n",
    "        txt_path = os.path.splitext(output_path)[0]+'.txt'\n",
    "        directory = os.path.dirname(output_path)\n",
    "        if not os.path.exists(directory):\n",
    "            os.makedirs(directory)\n",
    "        fid = open(txt_path, 'wb')\n",
    "        fid.write(input_path + ' ' + str(yaw) + ' ' + str(pitch) + ' ' + str(roll) + ' ' + str(face_x) + ' ' + str(face_y) + ' ' + str(face_w) + ' ' + str(face_h) + '\\n')\n",
    "        fid.close()\n",
    "        #Increasing the counter\n",
    "        counter = counter + 1 \n",
    "\n",
    "    #if the file does not exits it return an exception\n",
    "    else:\n",
    "        raise ValueError('Error: I cannot find the file specified: ' + str(input_path))\n",
    "\n",
    "#Once finished the iteration it closes the database\n",
    "c.close()\n",
    "print 'Done'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test\n"
     ]
    }
   ],
   "source": [
    "print 'test'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}

 practice/.ipynb_checkpoints/remove_KEPLER_test_split-checkpoint.ipynb

New file
@@ -0,0 +1,128 @@
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "AFLW2000 = '/Data/nruiz9/data/facial_landmarks/AFLW2000/'\n",
    "AFLW = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "list_2000 = '/Data/nruiz9/data/facial_landmarks/AFLW2000/filename_list_filtered.txt'\n",
    "list_normal = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped/filename_list.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1969\n"
     ]
    }
   ],
   "source": [
    "fid = open(list_2000, 'r')\n",
    "dict_2000 = dict()\n",
    "for line in fid:\n",
    "    line = line.strip('\\n').split('/')\n",
    "    dict_2000[line[-1]] = 1\n",
    "print len(dict_2000)\n",
    "fid.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1966 19109\n"
     ]
    }
   ],
   "source": [
    "fid = open(list_normal, 'r')\n",
    "naked = list_normal.strip('.txt')\n",
    "train = open(naked + '_train.txt', 'wb')\n",
    "test = open(naked + '_test.txt', 'wb')\n",
    "test_dict = dict()\n",
    "train_dict = dict()\n",
    "for line in fid:\n",
    "    line = line.strip('\\n')\n",
    "    name = line.split('/')[-1]\n",
    "    if name in dict_2000.keys():\n",
    "        test.write(line + '\\n')\n",
    "        test_dict[line] = 1\n",
    "    else:\n",
    "        train.write(line + '\\n')\n",
    "        train_dict[line] = 1\n",
    "\n",
    "print len(test_dict), len(train_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}

 practice/create_filtered_datasets.ipynb

@@ -96,19 +96,19 @@
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "AFLW = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped/'\n",
    "filenames = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped/filename_list_train.txt'"
    "AFLW = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose/'\n",
    "filenames = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose/filename_list.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
@@ -117,13 +117,13 @@
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "246\n"
      "243\n"
     ]
    }
   ],
   "source": [
    "fid = open(filenames, 'r')\n",
    "out = open(os.path.join(AFLW, 'filename_list_train_filtered.txt'), 'wb')\n",
    "out = open(os.path.join(AFLW, 'filename_list_filtered.txt'), 'wb')\n",
    "counter = 0\n",
    "for line in fid:\n",
    "    original_line = line\n",

 practice/load_AFLW-Copy1.ipynb

@@ -36,7 +36,7 @@
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
@@ -144,7 +144,7 @@
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },

 practice/load_AFLW-KEPLER_split.ipynb

New file
@@ -0,0 +1,190 @@
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import torch\n",
    "from torch.utils.serialization import load_lua\n",
    "import os\n",
    "import scipy.io as sio\n",
    "import cv2\n",
    "import math\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "import sqlite3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#Change this paths according to your directories\n",
    "images_path = \"/Data/nruiz9/data/facial_landmarks/AFLW/aflw/data/flickr/\"\n",
    "storing_path = \"/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose_test/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(573, 1)\n",
      "(232, 1)\n",
      "(165, 1)\n",
      "3\n",
      "(970, 1)\n"
     ]
    }
   ],
   "source": [
    "#Load KEPLER split file\n",
    "test_set = sio.loadmat('/Data/nruiz9/data/facial_landmarks/AFLW/testset.mat')\n",
    "print test_set['test'][0][0][0].shape\n",
    "print test_set['test'][0][0][1].shape\n",
    "print test_set['test'][0][0][2].shape\n",
    "print len(test_set['test'][0][0])\n",
    "\n",
    "id_test_set = np.concatenate([test_set['test'][0][0][0], test_set['test'][0][0][1], test_set['test'][0][0][2]])\n",
    "print id_test_set.shape\n",
    "\n",
    "# I will just use every single one for now. If results are not good I'll sample equally."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "970\n",
      "Done\n"
     ]
    }
   ],
   "source": [
    "#Image counter\n",
    "counter = 0\n",
    "\n",
    "#Open the sqlite database\n",
    "conn = sqlite3.connect('/Data/nruiz9/data/facial_landmarks/AFLW/aflw/data/aflw.sqlite')\n",
    "c = conn.cursor()\n",
    "\n",
    "#Creating the query string for retriving: roll, pitch, yaw and faces position\n",
    "#Change it according to what you want to retrieve\n",
    "select_string = \"faceimages.filepath, faces.face_id, facepose.roll, facepose.pitch, facepose.yaw, facerect.x, facerect.y, facerect.w, facerect.h\"\n",
    "from_string = \"faceimages, faces, facepose, facerect\"\n",
    "where_string = \"faces.face_id = facepose.face_id and faces.file_id = faceimages.file_id and faces.face_id = facerect.face_id\"\n",
    "query_string = \"SELECT \" + select_string + \" FROM \" + from_string + \" WHERE \" + where_string\n",
    "\n",
    "test_file_txt = '/Data/nruiz9/data/facial_landmarks/AFLW/KEPLER_test_split.txt'\n",
    "out_txt = open(test_file_txt, 'w')\n",
    "#It iterates through the rows returned from the query\n",
    "for row in c.execute(query_string):\n",
    "\n",
    "    #Using our specific query_string, the \"row\" variable will contain:\n",
    "    # row[0] = image path\n",
    "    # row[1] = face id\n",
    "    # row[2] = rollgma\n",
    "    # row[3] = pitch\n",
    "    # row[4] = yaw\n",
    "    # row[5] = face coord x\n",
    "    # row[6] = face coord y\n",
    "    # row[7] = face width\n",
    "    # row[8] = face heigh\n",
    "    if row[1] in id_test_set:\n",
    "        #Creating the full path names for input and output\n",
    "        input_path = images_path + str(row[0])\n",
    "        output_path = storing_path + str(row[0])\n",
    "\n",
    "        #If the file exist then open it       \n",
    "        if(os.path.isfile(input_path)  == True):\n",
    "\n",
    "            out_txt.write(str(row[0]) + '.jpg\\n')\n",
    "            #Increasing the counter\n",
    "            counter = counter + 1 \n",
    "\n",
    "        #if the file does not exits it return an exception\n",
    "        else:\n",
    "            raise ValueError('Error: I cannot find the file specified: ' + str(input_path))\n",
    "\n",
    "#Once finished the iteration it closes the database\n",
    "print counter\n",
    "out_txt.close()\n",
    "c.close()\n",
    "print 'Done'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test\n"
     ]
    }
   ],
   "source": [
    "print 'test'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}

 practice/remove_KEPLER_test_split.ipynb

New file
@@ -0,0 +1,128 @@
{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "AFLW2000 = '/Data/nruiz9/data/facial_landmarks/AFLW2000/'\n",
    "AFLW = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose/'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "list_2000 = '/Data/nruiz9/data/facial_landmarks/AFLW/KEPLER_test_split.txt'\n",
    "list_normal = '/Data/nruiz9/data/facial_landmarks/AFLW/aflw_cropped_loose/filename_list_filtered.txt'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "961\n"
     ]
    }
   ],
   "source": [
    "fid = open(list_2000, 'r')\n",
    "dict_2000 = dict()\n",
    "for line in fid:\n",
    "    line = line.strip('\\n').split('/')\n",
    "    dict_2000[line[-1]] = 1\n",
    "print len(dict_2000)\n",
    "fid.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "954 19842\n"
     ]
    }
   ],
   "source": [
    "fid = open(list_normal, 'r')\n",
    "naked = list_normal.strip('.txt')\n",
    "train = open(naked + '_train.txt', 'wb')\n",
    "test = open(naked + '_test.txt', 'wb')\n",
    "test_dict = dict()\n",
    "train_dict = dict()\n",
    "for line in fid:\n",
    "    line = line.strip('\\n')\n",
    "    name = line.split('/')[-1]\n",
    "    if name+'.jpg' in dict_2000.keys():\n",
    "        test.write(line + '\\n')\n",
    "        test_dict[line] = 1\n",
    "    else:\n",
    "        train.write(line + '\\n')\n",
    "        train_dict[line] = 1\n",
    "\n",
    "print len(test_dict), len(train_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda root]",
   "language": "python",
   "name": "conda-root-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}