segment_try.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n",
      "/data3/floraxue/cs282a/assn4env/lib/python3.5/site-packages/classification_models/resnext/__init__.py:4: UserWarning: Current ResNext models are deprecated, use keras.applications ResNeXt models\n",
      "  warnings.warn('Current ResNext models are deprecated, '\n"
     ]
    }
   ],
   "source": [
    "from skimage.io import imread, imsave, imshow\n",
    "import numpy as np\n",
    "from os.path import join\n",
    "from os import makedirs\n",
    "from segmentation_models import Unet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "filelist_x = glob.glob('/data3/floraxue/cs282a/data/dataset_1/raw_crop_image_100/*.tif')\n",
    "filelist_y = glob.glob('/data3/floraxue/cs282a/data/dataset_1/label_RGB_crop_100/*.tif')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "nn =10000\n",
    "X = np.array([imread(fname) for fname in filelist_x[:nn]])\n",
    "Y = np.array([imread(fname) for fname in filelist_y[:nn]])\n",
    "Y = np.sum(Y, axis=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "nn =10000\n",
    "X = np.reshape(X, [nn,256,256])\n",
    "Y = np.reshape(Y, [nn,256,256])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 256, 256)"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Collecting sklearn\n",
      "  Downloading https://files.pythonhosted.org/packages/1e/7a/dbb3be0ce9bd5c8b7e3d87328e79063f8b263b2b1bfa4774cb1147bfcd3f/sklearn-0.0.tar.gz\n",
      "Collecting scikit-learn (from sklearn)\n",
      "\u001b[?25l  Downloading https://files.pythonhosted.org/packages/38/7f/7d21bd89a97603a14d3770ca912d4402c394a8e50642f9b8b5cc6867a61a/scikit_learn-0.20.3-cp35-cp35m-manylinux1_x86_64.whl (5.3MB)\n",
      "\u001b[K     |████████████████████████████████| 5.3MB 13.7MB/s eta 0:00:01\n",
      "\u001b[?25hRequirement already satisfied: numpy>=1.8.2 in /data3/floraxue/cs282a/assn4env/lib/python3.5/site-packages (from scikit-learn->sklearn) (1.14.5)\n",
      "Requirement already satisfied: scipy>=0.13.3 in /data3/floraxue/cs282a/assn4env/lib/python3.5/site-packages (from scikit-learn->sklearn) (1.2.1)\n",
      "Building wheels for collected packages: sklearn\n",
      "  Building wheel for sklearn (setup.py) ... \u001b[?25ldone\n",
      "\u001b[?25h  Stored in directory: /home/floraxue/.cache/pip/wheels/76/03/bb/589d421d27431bcd2c6da284d5f2286c8e3b2ea3cf1594c074\n",
      "Successfully built sklearn\n",
      "Installing collected packages: scikit-learn, sklearn\n",
      "Successfully installed scikit-learn-0.20.3 sklearn-0.0\n"
     ]
    }
   ],
   "source": [
    "!pip install sklearn\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://github.com/qubvel/classification_models/releases/download/0.0.1/resnet34_imagenet_1000_no_top.h5\n",
      "85524480/85521592 [==============================] - 2s 0us/step\n"
     ]
    }
   ],
   "source": [
    "from segmentation_models import FPN\n",
    "from segmentation_models.backbones import get_preprocessing\n",
    "from segmentation_models.losses import bce_jaccard_loss\n",
    "from segmentation_models.metrics import iou_score\n",
    "\n",
    "\n",
    "BACKBONE = 'resnet34'\n",
    "preprocess_input = get_preprocessing(BACKBONE)\n",
    "\n",
    "# load your data\n",
    "\n",
    "x_train, x_val, y_train, y_val = train_test_split(X, Y, test_size=0.3, random_state=42)\n",
    "N = x_train.shape[-1]\n",
    "\n",
    "# preprocess input\n",
    "x_train = preprocess_input(x_train)\n",
    "x_val = preprocess_input(x_val)\n",
    "\n",
    "# define model\n",
    "base_model = FPN(BACKBONE, encoder_weights='imagenet')\n",
    "\n",
    "inp = Input(shape=(None, None, N))\n",
    "l1 = Conv2D(3, (1, 1))(inp) # map N channels data to 3 channels\n",
    "out = base_model(l1)\n",
    "\n",
    "model = Model(inp, out, name=base_model.name, classes=4, activation='softmax')\n",
    "\n",
    "#model = FPN(BACKBONE, input_shape=(None, 256, 256), classes=4, activation='softmax', encoder_weights=None)\n",
    "model.compile('Adam', loss='categorical_crossentropy', metrics=['categorical_accuracy'])\n",
    "\n",
    "# fit model\n",
    "model.fit(\n",
    "    x=x_train,\n",
    "    y=y_train,\n",
    "    batch_size=32,\n",
    "    epochs=2,\n",
    "    validation_data=(x_val, y_val),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x_val.shape"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}