predict_TF.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Make a prediction on CO-tip images from folder\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import glob\n",
    "import os\n",
    "from PIL import Image\n",
    "from keras import backend as K\n",
    "from keras.models import load_model\n",
    "from utils import *\n",
    "\n",
    "\n",
    "# model parameters\n",
    "crop_size      = 16  # pixels size of images to feed in input of CNN\n",
    "model_dir      = './pretrained_weights/' \n",
    "\n",
    "K.set_image_dim_ordering('th')\n",
    "if K.image_data_format() == 'channels_first':\n",
    "    input_shape = (1,crop_size, crop_size)\n",
    "else:\n",
    "    input_shape = (crop_size, crop_size,1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### load data to predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dataX: (36, 1, 16, 16)\n"
     ]
    }
   ],
   "source": [
    "#load data to predict\n",
    "path_to_data  = './databaseNew6/test2/bads'   # <======= path to folder with CO images\n",
    "path_to_preds = './'  # output prediction will be plotted here to predicted.png \n",
    "dataX = []\n",
    "file_names = []\n",
    "images = sorted(glob.glob(path_to_data+'/'+'*.png'))\n",
    "for image_path in images:\n",
    "    _, file_name = os.path.split(image_path)\n",
    "    image = np.array((Image.open(image_path).resize((crop_size,crop_size), Image.ANTIALIAS))).astype(np.float32)\n",
    "    dataX.append(np.flipud(image))\n",
    "    file_names.append(file_name)\n",
    "dataX = np.expand_dims(np.array(dataX), axis = 1)\n",
    "print (f'dataX: {dataX.shape}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Model loaded from ./pretrained_weights/model.h5\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_1 (Conv2D)            (None, 4, 14, 14)         40        \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_1 (LeakyReLU)    (None, 4, 14, 14)         0         \n",
      "_________________________________________________________________\n",
      "spatial_dropout2d_1 (Spatial (None, 4, 14, 14)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_2 (Conv2D)            (None, 4, 12, 12)         148       \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_2 (LeakyReLU)    (None, 4, 12, 12)         0         \n",
      "_________________________________________________________________\n",
      "spatial_dropout2d_2 (Spatial (None, 4, 12, 12)         0         \n",
      "_________________________________________________________________\n",
      "average_pooling2d_1 (Average (None, 4, 6, 6)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_3 (Conv2D)            (None, 8, 4, 4)           296       \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_3 (LeakyReLU)    (None, 8, 4, 4)           0         \n",
      "_________________________________________________________________\n",
      "spatial_dropout2d_3 (Spatial (None, 8, 4, 4)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_4 (Conv2D)            (None, 8, 2, 2)           584       \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_4 (LeakyReLU)    (None, 8, 2, 2)           0         \n",
      "_________________________________________________________________\n",
      "spatial_dropout2d_4 (Spatial (None, 8, 2, 2)           0         \n",
      "_________________________________________________________________\n",
      "flatten_1 (Flatten)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 32)                1056      \n",
      "_________________________________________________________________\n",
      "leaky_re_lu_5 (LeakyReLU)    (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 32)                0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 1)                 33        \n",
      "_________________________________________________________________\n",
      "activation_1 (Activation)    (None, 1)                 0         \n",
      "=================================================================\n",
      "Total params: 2,157\n",
      "Trainable params: 2,157\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model_path = model_dir + 'model.h5'\n",
    "\n",
    "# Load model\n",
    "model = load_model(model_path)\n",
    "print(f'\\nModel loaded from {model_path}')\n",
    "# summarize model.\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "add_norm_CO(dataX)\n",
    "\n",
    "outputs = model.predict_on_batch(dataX)\n",
    "probs = np.squeeze(np.array(outputs))\n",
    "preds = np.squeeze(np.round(outputs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_preds_with_title(dataX,probs,file_names, cmap=cm.gray):\n",
    "    cols = 10\n",
    "    rows = -(-len(file_names) // cols)\n",
    "    fig = plt.figure(figsize=(3.0*cols,3.0*rows))\n",
    "    \n",
    "    for i in range(len(file_names)):\n",
    "        sp = fig.add_subplot(rows,cols,i+1)#, origin=\"lower\"\n",
    "        sp.imshow(dataX[i,0,:,:], cmap = cmap)   \n",
    "        sp.axis('Off')\n",
    "        sp.set_title(f'pred = {probs[i]: 0.3f}\\n{file_names[i]}') \n",
    "    \n",
    "    save_name = 'predicted.png'\n",
    "    plt.savefig('./'+save_name, bbox_inches='tight', dpi=200)\n",
    "    plt.show()\n",
    "    plt.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 2160x864 with 36 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot prediction results to image\n",
    "plot_preds_with_title(dataX,probs,file_names)\n"
   ]
  },
  {
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}