test.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/ubuntu/anaconda3/envs/tensorflow_p36/lib/python3.6/importlib/_bootstrap.py:219: RuntimeWarning: compiletime version 3.5 of module 'tensorflow.python.framework.fast_tensor_util' does not match runtime version 3.6\n",
      "  return f(*args, **kwds)\n"
     ]
    }
   ],
   "source": [
    "import CNN2Head_input\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import BKNetStyle\n",
    "from const import *\n",
    "import cv2\n",
    "\n",
    "from IPython.display import Image, display"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "sess = tf.InteractiveSession()\n",
    "x = tf.placeholder(tf.float32, [None, 48, 48, 1])\n",
    "\n",
    "y_smile_conv, y_gender_conv, y_age_conv, phase_train, keep_prob = BKNetStyle.BKNetModel(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Restore model\n",
      "INFO:tensorflow:Restoring parameters from ./save/current/model.ckpt\n",
      "OK\n"
     ]
    },
    {
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {
      "image/jpeg": {
       "height": 100,
       "width": 100
      }
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Smile: Yes [[ 0.00668042  0.99331963]]\n",
      "Gender: Female [[ 0.70619285  0.29380712]]\n",
      "Age: Middle (30 <= age <45) [[ 0.34444451  0.44706547  0.17603277  0.0324572 ]]\n"
     ]
    }
   ],
   "source": [
    "PATH = '/home/ubuntu/coding/cnn/datasets/genki4k/files/file0009.jpg'\n",
    "img = cv2.imread(PATH)\n",
    "img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
    "img = cv2.resize(img,(48, 48))\n",
    "\n",
    "img = (img - 128) / 255.0\n",
    "T = np.zeros([48, 48, 1])\n",
    "T[:, :, 0] = img\n",
    "test_img = []\n",
    "test_img.append(T)\n",
    "test_img = np.asarray(test_img)\n",
    "\n",
    "print('Restore model')\n",
    "saver = tf.train.Saver()\n",
    "saver.restore(sess, SAVE_FOLDER + 'model.ckpt')\n",
    "print('OK')\n",
    "\n",
    "display(Image(filename = PATH, width=100, height=100))\n",
    "\n",
    "predict_y_smile_conv = sess.run(y_smile_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "predict_y_gender_conv = sess.run(y_gender_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "predict_y_age_conv = sess.run(y_age_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "\n",
    "smile_label = \"No \" if np.argmax(predict_y_smile_conv)==0 else \"Yes \"\n",
    "gender_label = \"Female \" if np.argmax(predict_y_gender_conv)==0 else \"Male \"\n",
    "argmax_predict_age = np.argmax(predict_y_age_conv)\n",
    "if argmax_predict_age==0:\n",
    "    age_label = \"Young (age < 30yrs) \"\n",
    "elif argmax_predict_age==1:\n",
    "    age_label = \"Middle (30 <= age <45) \"\n",
    "elif argmax_predict_age==2:\n",
    "    age_label = \"Old (45 <= age < 60) \"\n",
    "else :\n",
    "    age_label = \"Very Old (60 <= age) \"\n",
    "\n",
    "print('Smile: ' + smile_label + str(predict_y_smile_conv))\n",
    "print('Gender: ' + gender_label + str(predict_y_gender_conv))\n",
    "print('Age: ' + age_label + str(predict_y_age_conv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Restore model\n",
      "INFO:tensorflow:Restoring parameters from ./save/current/model.ckpt\n",
      "OK\n"
     ]
    },
    {
     "data": {
      "image/jpeg": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "metadata": {
      "image/jpeg": {
       "height": 100,
       "width": 100
      }
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Smile: Yes [[ 0.00149914  0.99850082]]\n",
      "Gender: Male [[ 0.24222387  0.75777608]]\n",
      "Age: Middle (30 <= age <45) [[ 0.11998766  0.42773849  0.35627618  0.09599768]]\n"
     ]
    }
   ],
   "source": [
    "PATH = '/home/ubuntu/coding/cnn/datasets/imdb_crop/00/nm0000100_rm3653409024_1955-1-6_2003.jpg'\n",
    "img = cv2.imread(PATH)\n",
    "img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)\n",
    "img = cv2.resize(img,(48, 48))\n",
    "\n",
    "img = (img - 128) / 255.0\n",
    "T = np.zeros([48, 48, 1])\n",
    "T[:, :, 0] = img\n",
    "test_img = []\n",
    "test_img.append(T)\n",
    "test_img = np.asarray(test_img)\n",
    "\n",
    "print('Restore model')\n",
    "saver = tf.train.Saver()\n",
    "saver.restore(sess, SAVE_FOLDER + 'model.ckpt')\n",
    "print('OK')\n",
    "\n",
    "display(Image(filename = PATH, width=100, height=100))\n",
    "\n",
    "predict_y_smile_conv = sess.run(y_smile_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "predict_y_gender_conv = sess.run(y_gender_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "predict_y_age_conv = sess.run(y_age_conv, feed_dict={x: test_img, phase_train: False, keep_prob: 1})\n",
    "\n",
    "smile_label = \"No \" if np.argmax(predict_y_smile_conv)==0 else \"Yes \"\n",
    "gender_label = \"Female \" if np.argmax(predict_y_gender_conv)==0 else \"Male \"\n",
    "argmax_predict_age = np.argmax(predict_y_age_conv)\n",
    "if argmax_predict_age==0:\n",
    "    age_label = \"Young (age < 30yrs) \"\n",
    "elif argmax_predict_age==1:\n",
    "    age_label = \"Middle (30 <= age <45) \"\n",
    "elif argmax_predict_age==2:\n",
    "    age_label = \"Old (45 <= age < 60) \"\n",
    "else :\n",
    "    age_label = \"Very Old (60 <= age) \"\n",
    "\n",
    "print('Smile: ' + smile_label + str(predict_y_smile_conv))\n",
    "print('Gender: ' + gender_label + str(predict_y_gender_conv))\n",
    "print('Age: ' + age_label + str(predict_y_age_conv))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}