Example.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fpm import *\n",
    "from params import *\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import torch\n",
    "import scipy.io as io\n",
    "import cv2\n",
    "\n",
    "from utils import get_label\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "models = {\n",
    "    '64_raw': ('models/frcnn_64_raw', False),\n",
    "    '64_pre': ('models/frcnn_64_pre', True),\n",
    "    '128_raw': ('models/frcnn_128_raw', False),\n",
    "    '128_pre': ('models/frcnn_128_pre', True),\n",
    "    '256_raw': ('models/frcnn_256_raw', False),\n",
    "    '256_pre': ('models/frcnn_256_pre', True),\n",
    "    '512_raw': ('models/frcnn_512_raw', False),\n",
    "    '512_pre': ('models/frcnn_512_pre', True),\n",
    "    'multisize_raw': ('models/frcnn_multisize_raw', False),\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = '64_raw'  # ['64_raw', '64_pre', '128_raw', '128_pre', '256_raw', '256_pre', '512_raw', '512_pre', 'multisize_raw']\n",
    "\n",
    "cfg_path, preprocess_fft = models[model]\n",
    "cfg = pickle.load(open(os.path.join(cfg_path, 'cfg.pkl'), 'rb'))\n",
    "cfg.MODEL.WEIGHTS = os.path.join(cfg_path, 'model_final.pth')\n",
    "predictor = Predictor(cfg, preprocess_fft=preprocess_fft)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Instances(num_instances=1, image_height=128, image_width=128, fields=[pred_boxes: Boxes(tensor([[  0.0000,   3.9708, 125.3443, 127.8070]], device='cuda:0',\n",
       "       grad_fn=<IndexBackward>)), scores: tensor([0.4509], device='cuda:0', grad_fn=<IndexBackward>), pred_classes: tensor([0], device='cuda:0')])"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tiff_path = 'tiffs/4.2.03_128.tiff'\n",
    "img = read_tiff(tiff_path)[0]\n",
    "img = torch.from_numpy(img).cuda()\n",
    "img = [\n",
    "    {\n",
    "        'image': img,\n",
    "    }\n",
    "]\n",
    "\n",
    "predictor.model(img)[0]['instances']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### An example tun-through of the available functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading the model..."
     ]
    }
   ],
   "source": [
    "tiff_path = 'tiffs/4.2.03_128.tiff'\n",
    "df = pd.read_csv('~/Eval-FPM/data_512/labels.csv')\n",
    "row = get_label(f'data_512/images/4.2.03.tiff', df, return_row=True)\n",
    "illumination_params['model'] = '128_raw'  # ['64_raw', '64_pre', '128_raw', '128_pre', '256_raw', '256_pre', '512_raw', '512_pre', 'multisize_raw']\n",
    "reconstruction_params['scale'] = 2\n",
    "illumination_params['calibrate'] = False\n",
    "\n",
    "# Illumination Estimation\n",
    "discs, radii = get_illumination(tiff_path, **illumination_params)\n",
    "\n",
    "# Save Illumination Estimation Results\n",
    "save_illumination(discs, radii, tiff_path, save_params)\n",
    "\n",
    "# Reconstruction\n",
    "obj, pupil, imgs = get_reconstruction(tiff_path, discs, row, reconstruction_params)\n",
    "\n",
    "# Save Reconstruction Results\n",
    "save_reconstruction(obj, pupil, imgs, tiff_path, save_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/suyog_jadhav1_gmail_com/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:4: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  after removing the cwd from sys.path.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading the model...Done!\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Processing Apodization', max=60.0, style=ProgressStyle(de…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Processing PSD', max=60.0, style=ProgressStyle(descriptio…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "77d3f9931fb74b77b8378a0f2fdb58fe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Running Inference', max=60.0, style=ProgressStyle(descrip…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Processing Apodization', max=60.0, style=ProgressStyle(de…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performing Reconstruction..."
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, description='Working...', max=40.0, style=ProgressStyle(description_wi…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done!\n"
     ]
    }
   ],
   "source": [
    "tiff_path = '4.2.03_64.tiff'\n",
    "df = pd.read_csv('~/Eval-FPM/data_512/labels.csv')\n",
    "row = get_label(f'data_512/images/4.2.03.tiff', df, return_row=True)\n",
    "row['IMAGESIZE'] = 64\n",
    "illumination_params['do_psd'] = True\n",
    "illumination_params['visualize'] = False\n",
    "illumination_params['model'] = '64_pre'\n",
    "\n",
    "reconstruction_params['do_psd'] = False\n",
    "reconstruction_params['window'] = 'tukey'\n",
    "reconstruction_params['a'] = 0.25\n",
    "reconstruction_params['n_iters'] = 40\n",
    "\n",
    "# Illumination Estimation\n",
    "discs, radii = get_illumination(tiff_path, **illumination_params)\n",
    "\n",
    "# Save Illumination Estimation Results\n",
    "# save_illumination(discs, radii, tiff_path, save_params)\n",
    "\n",
    "# Reconstruction\n",
    "obj, pupil, imgs = get_reconstruction(tiff_path, discs, row, reconstruction_params)\n",
    "\n",
    "# Save Reconstruction Results\n",
    "# save_reconstruction(obj, pupil, imgs, tiff_path, save_params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 106,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f3c9a0737d0>"
      ]
     },
     "execution_count": 106,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(cp.asnumpy(cp.abs(obj)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 107,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Saving illumination results...Done!\n",
      "Saving reconstruction results...Done!\n"
     ]
    }
   ],
   "source": [
    "# Save Illumination Estimation Results\n",
    "save_illumination(discs, radii, tiff_path, save_params)\n",
    "\n",
    "# Save Reconstruction Results\n",
    "save_reconstruction(obj, pupil, imgs, tiff_path, save_params)"
   ]
  },
  {
   "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}