data_eda.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import json\n",
    "import cv2\n",
    "from PIL import Image\n",
    "import numpy as np\n",
    "from tqdm import tqdm\n",
    "import albumentations as A\n",
    "import torch\n",
    "from torch.utils.data import Dataset\n",
    "from sklearn.model_selection import GroupKFold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "IMAGE_ROOT = \"/opt/ml/input/data/train/DCM\"\n",
    "LABEL_ROOT = \"/opt/ml/input/data/train/outputs_json\"\n",
    "\n",
    "CLASSES = [\n",
    "    'finger-1', 'finger-2', 'finger-3', 'finger-4', 'finger-5',\n",
    "    'finger-6', 'finger-7', 'finger-8', 'finger-9', 'finger-10',\n",
    "    'finger-11', 'finger-12', 'finger-13', 'finger-14', 'finger-15',\n",
    "    'finger-16', 'finger-17', 'finger-18', 'finger-19', 'Trapezium',\n",
    "    'Trapezoid', 'Capitate', 'Hamate', 'Scaphoid', 'Lunate',\n",
    "    'Triquetrum', 'Pisiform', 'Radius', 'Ulna',\n",
    "]\n",
    "\n",
    "CLASS2IND = {v: i for i, v in enumerate(CLASSES)}\n",
    "IND2CLASS = {v: k for k, v in CLASS2IND.items()}\n",
    "\n",
    "pngs = {\n",
    "    os.path.relpath(os.path.join(root, fname), start=IMAGE_ROOT)\n",
    "    for root, _dirs, files in os.walk(IMAGE_ROOT)\n",
    "    for fname in files\n",
    "    if os.path.splitext(fname)[1].lower() == \".png\"\n",
    "}\n",
    "pngs = sorted(pngs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Calculate Image Mean & Std...: 100%|██████████| 800/800 [04:10<00:00,  3.19it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Global Mean = [0.12099939 0.12099939 0.12099939]\n",
      "Global Std = [0.0363468 0.0363468 0.0363468]\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "temp_mean, temp_std = [], []\n",
    "\n",
    "for png in tqdm(pngs, desc=\"Calculate Image Mean & Std...\"):\n",
    "    image_path = os.path.join(IMAGE_ROOT, png)\n",
    "    image = cv2.imread(image_path)\n",
    "\n",
    "    per_mean = image.mean(axis=(0, 1))\n",
    "    per_std = (image ** 2).mean(axis=(0, 1))\n",
    "    temp_mean.append(per_mean)\n",
    "    temp_std.append(per_std)\n",
    "\n",
    "print()\n",
    "global_mean = (np.mean(temp_mean, axis=0)) / 255\n",
    "global_std = (np.mean(temp_std, axis=0) - global_mean ** 2) ** 0.5 / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Global Mean = [0.12099939 0.12099939 0.12099939]\n",
      "Global Std = [0.0363468 0.0363468 0.0363468]\n"
     ]
    }
   ],
   "source": [
    "print(f\"Global Mean = {global_mean}\")\n",
    "print(f\"Global Std = {global_std}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Mean & Std Calculation for Albumentations.Normalize\n",
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.8.5"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}