cuneiform-sign-detection-code/experiments/line_segmentation/precompute_line_segmentations.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pre-compute and store line segmentations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from PIL import Image\n",
    "from ast import literal_eval"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# torch\n",
    "import torch\n",
    "import torchvision\n",
    "# addons\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#%pylab inline\n",
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# for auto-reloading external modules\n",
    "# see http://stackoverflow.com/questions/1907993/autoreload-of-modules-in-ipython\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "relative_path = '../../'\n",
    "# ensure that parent path is on the python path in order to have all packages available\n",
    "import sys, os\n",
    "parent_path = os.path.join(os.getcwd(), relative_path)\n",
    "parent_path = os.path.realpath(parent_path)  # os.path.abspath(...)\n",
    "sys.path.insert(0, parent_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from lib.models.trained_model_loader import get_line_net_fcn\n",
    "from lib.datasets.cunei_dataset_segments import CuneiformSegments, get_segment_meta\n",
    "from lib.transliteration.sign_labels import get_label_list\n",
    "from lib.utils.transform_utils import UnNormalize\n",
    "\n",
    "from lib.detection.run_gen_line_detection import gen_line_detections"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Config Basics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# toggle generation\n",
    "save_line_detections = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# set line segmentation network\n",
    "line_model_version = 'v002'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# dataset config\n",
    "collections = ['test', 'train', 'saa01',  'saa05', 'saa06', 'saa08', 'saa09', 'saa10', 'saa13', 'saa16'] \n",
    "#collections = ['saa01', 'saa05']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "data_layer_params = dict(batch_size=[128, 16],\n",
    "                         img_channels=1,\n",
    "                         gray_mean=[0.5],\n",
    "                         gray_std=[1.0], \n",
    "                         num_classes = 2\n",
    "                         )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Config Data Augmentation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "num_classes = data_layer_params['num_classes']\n",
    "num_c = data_layer_params['img_channels']\n",
    "gray_mean = data_layer_params['gray_mean']\n",
    "gray_std = data_layer_params['gray_std']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "re_transform = torchvision.transforms.Compose([\n",
    "    UnNormalize(mean=gray_mean, std=gray_std),\n",
    "    torchvision.transforms.ToPILImage(),\n",
    "                                              ])\n",
    "re_transform_rgb = torchvision.transforms.Compose([\n",
    "    UnNormalize(mean=gray_mean * 3, std=gray_std * 3),\n",
    "    torchvision.transforms.ToPILImage(),\n",
    "                                              ])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#use_gpu = torch.cuda.is_available()\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "model_fcn = get_line_net_fcn(line_model_version, device, num_classes=num_classes, num_c=num_c)\n",
    "print(model_fcn)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "for saa_version in collections:\n",
    "    print('collection: <><>{}<><>'.format(saa_version))\n",
    "    \n",
    "    ### Get collection dataset\n",
    "    dataset = CuneiformSegments(collections=[saa_version], relative_path=relative_path, \n",
    "                                only_annotated=False, only_assigned=True, preload_segments=False)\n",
    "    \n",
    "    # filter collection dataset - OPTIONAL\n",
    "    didx_list = range(len(dataset))\n",
    "    \n",
    "    ### Generate line detections\n",
    "    gen_line_detections(didx_list, dataset, saa_version, relative_path,\n",
    "                        line_model_version, model_fcn, re_transform, device,\n",
    "                        save_line_detections)     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}