Analysis.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import matplotlib.pyplot as plt\n",
    "import cv2\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Detect if we have a GPU available\n",
    "if torch.backends.mps.is_available() and torch.backends.mps.is_built():\n",
    "    device = torch.device(\"mps\")\n",
    "elif torch.cuda.is_available():\n",
    "    device = torch.device(\"cuda:0\")\n",
    "else:\n",
    "    device = torch.device(\"cpu\")\n",
    "\n",
    "print(f\"Using device: {device}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DeepLabV3(\n",
       "  (backbone): IntermediateLayerGetter(\n",
       "    (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
       "    (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (relu): ReLU(inplace=True)\n",
       "    (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
       "    (layer1): Sequential(\n",
       "      (0): Bottleneck(\n",
       "        (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): Bottleneck(\n",
       "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (2): Bottleneck(\n",
       "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "    )\n",
       "    (layer2): Sequential(\n",
       "      (0): Bottleneck(\n",
       "        (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
       "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): Bottleneck(\n",
       "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (2): Bottleneck(\n",
       "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (3): Bottleneck(\n",
       "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "    )\n",
       "    (layer3): Sequential(\n",
       "      (0): Bottleneck(\n",
       "        (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (2): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (3): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (4): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (5): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (6): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (7): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (8): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (9): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (10): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (11): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (12): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (13): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (14): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (15): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (16): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (17): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (18): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (19): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (20): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (21): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (22): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "    )\n",
       "    (layer4): Sequential(\n",
       "      (0): Bottleneck(\n",
       "        (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "        (downsample): Sequential(\n",
       "          (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        )\n",
       "      )\n",
       "      (1): Bottleneck(\n",
       "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(4, 4), dilation=(4, 4), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "      (2): Bottleneck(\n",
       "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(4, 4), dilation=(4, 4), bias=False)\n",
       "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (relu): ReLU(inplace=True)\n",
       "      )\n",
       "    )\n",
       "  )\n",
       "  (classifier): DeepLabHead(\n",
       "    (0): ASPP(\n",
       "      (convs): ModuleList(\n",
       "        (0): Sequential(\n",
       "          (0): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU()\n",
       "        )\n",
       "        (1): ASPPConv(\n",
       "          (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(12, 12), dilation=(12, 12), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU()\n",
       "        )\n",
       "        (2): ASPPConv(\n",
       "          (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(24, 24), dilation=(24, 24), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU()\n",
       "        )\n",
       "        (3): ASPPConv(\n",
       "          (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(36, 36), dilation=(36, 36), bias=False)\n",
       "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (2): ReLU()\n",
       "        )\n",
       "        (4): ASPPPooling(\n",
       "          (0): AdaptiveAvgPool2d(output_size=1)\n",
       "          (1): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "          (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "          (3): ReLU()\n",
       "        )\n",
       "      )\n",
       "      (project): Sequential(\n",
       "        (0): Conv2d(1280, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
       "        (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "        (2): ReLU()\n",
       "        (3): Dropout(p=0.5, inplace=False)\n",
       "      )\n",
       "    )\n",
       "    (1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "    (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (3): ReLU()\n",
       "    (4): Conv2d(256, 1, kernel_size=(1, 1), stride=(1, 1))\n",
       "  )\n",
       "  (aux_classifier): FCNHead(\n",
       "    (0): Conv2d(1024, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
       "    (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
       "    (2): ReLU()\n",
       "    (3): Dropout(p=0.1, inplace=False)\n",
       "    (4): Conv2d(256, 21, kernel_size=(1, 1), stride=(1, 1))\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Load the trained model \n",
    "model = torch.load('./CFExp/weights.pt')\n",
    "# Set the model to evaluate mode\n",
    "model.eval()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Read the log file using pandas into a dataframe\n",
    "df = pd.read_csv('./CFExp/log.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Training and testing loss, f1_score and auroc values for the model trained on the CrackForest dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot all the values with respect to the epochs\n",
    "df.plot(x='epoch',figsize=(15,8));"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train_auroc    0.936169\n",
      "Test_auroc     0.841937\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(df[['Train_auroc','Test_auroc']].max())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Sample Prediction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "ino = 2\n",
    "# Read  a sample image and mask from the data-set\n",
    "img = cv2.imread(f'./CrackForest/Images/{ino:03d}.jpg').transpose(2,0,1).reshape(1,3,320,480)\n",
    "mask = cv2.imread(f'./CrackForest/Masks/{ino:03d}_label.PNG')\n",
    "with torch.no_grad():\n",
    "    a = model(torch.from_numpy(img).type(torch.cuda.FloatTensor)/255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([2.64900e+03, 1.31774e+05, 6.76700e+03, 4.42500e+03, 3.60900e+03,\n",
       "        2.36500e+03, 1.32800e+03, 6.20000e+02, 4.80000e+01, 1.50000e+01]),\n",
       " array([-0.2715176 , -0.11536801,  0.04078159,  0.1969312 ,  0.35308078,\n",
       "         0.5092304 ,  0.66538   ,  0.82152957,  0.9776792 ,  1.1338288 ,\n",
       "         1.2899784 ], dtype=float32),\n",
       " <BarContainer object of 10 artists>)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAD4CAYAAAAZ1BptAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAAUZUlEQVR4nO3dfZBd9X3f8fcnUsF2GhCYLSESqchYbivTJMYaTOpO4lguCJyxmCl2YeKguBprHOM0fZhJRP0HHdvMmD6EhhlMqwEVwTgGQp2iqUWpDHiYJhFmCS4gCLAB24jwsEECt2VsLOfbP+5P7u2yP620V7q7wPs1c2fP+Z7fOed772r3s+fhXqWqkCRpNj+20A1IkhYvQ0KS1GVISJK6DAlJUpchIUnqWrrQDRxpJ510Uq1cuXKh25Ck15X777//L6tqYmb9DRcSK1euZHJycqHbkKTXlSTfnq3u6SZJUpchIUnqMiQkSV2GhCSpy5CQJHUZEpKkLkNCktRlSEiSugwJSVLXG+4d169XKzd/dUH2+60vfGhB9ivp9cEjCUlSlyEhSeoyJCRJXYaEJKnLkJAkdRkSkqQuQ0KS1GVISJK6DAlJUpchIUnqMiQkSV2GhCSpy5CQJHUZEpKkrjlDIsnWJC8keXio9m+S/FmSB5P8YZJlQ8suTTKV5LEk5wzV17XaVJLNQ/XTktzb6jcnOabVj23zU235yiP1pCVJh+ZQjiSuB9bNqO0ETq+qnwUeBy4FSLIauBB4V1vni0mWJFkCXA2cC6wGLmpjAa4ArqyqdwD7gI2tvhHY1+pXtnGSpDGaMySq6h5g74zaf6+q/W12F7CiTa8Hbqqq71fVU8AUcGZ7TFXVk1X1KnATsD5JgA8At7b1twHnD21rW5u+FVjbxkuSxuRIXJP4x8DtbXo58PTQsj2t1qu/HXhpKHAO1P+/bbXlL7fxr5FkU5LJJJPT09MjPyFJ0sBIIZHkM8B+4EtHpp35qaotVbWmqtZMTEwsZCuS9IYy7//jOsmvA78CrK2qauVngFOHhq1oNTr1F4FlSZa2o4Xh8Qe2tSfJUuD4Nl6SNCbzOpJIsg74beDDVfXK0KLtwIXtzqTTgFXAN4D7gFXtTqZjGFzc3t7C5W7ggrb+BuC2oW1taNMXAHcNhZEkaQzmPJJI8mXg/cBJSfYAlzG4m+lYYGe7lryrqj5ZVbuT3AI8wuA01CVV9cO2nU8DdwBLgK1Vtbvt4neAm5J8HngAuK7VrwNuTDLF4ML5hUfg+UqSDsOcIVFVF81Svm6W2oHxlwOXz1LfAeyYpf4kg7ufZta/B3xkrv4kSUeP77iWJHUZEpKkLkNCktRlSEiSugwJSVKXISFJ6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkroMCUlSlyEhSeoyJCRJXYaEJKnLkJAkdRkSkqQuQ0KS1GVISJK6DAlJUpchIUnqmjMkkmxN8kKSh4dqJybZmeSJ9vWEVk+Sq5JMJXkwyRlD62xo459IsmGo/p4kD7V1rkqSg+1DkjQ+h3IkcT2wbkZtM3BnVa0C7mzzAOcCq9pjE3ANDH7hA5cB7wXOBC4b+qV/DfCJofXWzbEPSdKYzBkSVXUPsHdGeT2wrU1vA84fqt9QA7uAZUlOAc4BdlbV3qraB+wE1rVlx1XVrqoq4IYZ25ptH5KkMZnvNYmTq+rZNv0ccHKbXg48PTRuT6sdrL5nlvrB9iFJGpORL1y3I4A6Ar3Mex9JNiWZTDI5PT19NFuRpDeV+YbE8+1UEe3rC63+DHDq0LgVrXaw+opZ6gfbx2tU1ZaqWlNVayYmJub5lCRJM803JLYDB+5Q2gDcNlS/uN3ldBbwcjtldAdwdpIT2gXrs4E72rLvJjmr3dV08YxtzbYPSdKYLJ1rQJIvA+8HTkqyh8FdSl8AbkmyEfg28NE2fAdwHjAFvAJ8HKCq9ib5HHBfG/fZqjpwMfxTDO6geitwe3twkH1IksZkzpCoqos6i9bOMraASzrb2QpsnaU+CZw+S/3F2fYhSRof33EtSeoyJCRJXYaEJKnLkJAkdRkSkqQuQ0KS1GVISJK6DAlJUpchIUnqMiQkSV2GhCSpy5CQJHUZEpKkLkNCktRlSEiSugwJSVKXISFJ6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkrpGCokk/yzJ7iQPJ/lykrckOS3JvUmmktyc5Jg29tg2P9WWrxzazqWt/liSc4bq61ptKsnmUXqVJB2+eYdEkuXAPwHWVNXpwBLgQuAK4MqqegewD9jYVtkI7Gv1K9s4kqxu670LWAd8McmSJEuAq4FzgdXARW2sJGlMRj3dtBR4a5KlwNuAZ4EPALe25duA89v0+jZPW742SVr9pqr6flU9BUwBZ7bHVFU9WVWvAje1sZKkMZl3SFTVM8C/Bb7DIBxeBu4HXqqq/W3YHmB5m14OPN3W3d/Gv324PmOdXv01kmxKMplkcnp6er5PSZI0wyinm05g8Jf9acBPAT/O4HTR2FXVlqpaU1VrJiYmFqIFSXpDGuV00weBp6pquqp+AHwFeB+wrJ1+AlgBPNOmnwFOBWjLjwdeHK7PWKdXlySNySgh8R3grCRva9cW1gKPAHcDF7QxG4Db2vT2Nk9bfldVVatf2O5+Og1YBXwDuA9Y1e6WOobBxe3tI/QrSTpMS+ceMruqujfJrcCfAvuBB4AtwFeBm5J8vtWua6tcB9yYZArYy+CXPlW1O8ktDAJmP3BJVf0QIMmngTsY3Dm1tap2z7dfSdLhm3dIAFTVZcBlM8pPMrgzaebY7wEf6WzncuDyWeo7gB2j9ChJmj/fcS1J6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkroMCUlSlyEhSeoyJCRJXYaEJKnLkJAkdRkSkqQuQ0KS1GVISJK6DAlJUpchIUnqMiQkSV2GhCSpy5CQJHUZEpKkLkNCktQ1UkgkWZbk1iR/luTRJL+Q5MQkO5M80b6e0MYmyVVJppI8mOSMoe1saOOfSLJhqP6eJA+1da5KklH6lSQdnlGPJH4P+G9V9beBnwMeBTYDd1bVKuDONg9wLrCqPTYB1wAkORG4DHgvcCZw2YFgaWM+MbTeuhH7lSQdhnmHRJLjgV8ErgOoqler6iVgPbCtDdsGnN+m1wM31MAuYFmSU4BzgJ1Vtbeq9gE7gXVt2XFVtauqCrhhaFuSpDEY5UjiNGAa+E9JHkhybZIfB06uqmfbmOeAk9v0cuDpofX3tNrB6ntmqUuSxmSUkFgKnAFcU1XvBv4P/+/UEgDtCKBG2MchSbIpyWSSyenp6aO9O0l60xglJPYAe6rq3jZ/K4PQeL6dKqJ9faEtfwY4dWj9Fa12sPqKWeqvUVVbqmpNVa2ZmJgY4SlJkobNOySq6jng6SR/q5XWAo8A24EDdyhtAG5r09uBi9tdTmcBL7fTUncAZyc5oV2wPhu4oy37bpKz2l1NFw9tS5I0BktHXP83gS8lOQZ4Evg4g+C5JclG4NvAR9vYHcB5wBTwShtLVe1N8jngvjbus1W1t01/CrgeeCtwe3tIksZkpJCoqm8Ca2ZZtHaWsQVc0tnOVmDrLPVJ4PRRepQkzZ/vuJYkdRkSkqQuQ0KS1GVISJK6DAlJUpchIUnqMiQkSV2GhCSpy5CQJHUZEpKkLkNCktRlSEiSugwJSVKXISFJ6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkroMCUlSlyEhSeoyJCRJXSOHRJIlSR5I8l/b/GlJ7k0yleTmJMe0+rFtfqotXzm0jUtb/bEk5wzV17XaVJLNo/YqSTo8R+JI4reAR4fmrwCurKp3APuAja2+EdjX6le2cSRZDVwIvAtYB3yxBc8S4GrgXGA1cFEbK0kak5FCIskK4EPAtW0+wAeAW9uQbcD5bXp9m6ctX9vGrwduqqrvV9VTwBRwZntMVdWTVfUqcFMbK0kak1GPJP498NvAX7X5twMvVdX+Nr8HWN6mlwNPA7TlL7fxP6rPWKdXf40km5JMJpmcnp4e8SlJkg6Yd0gk+RXghaq6/wj2My9VtaWq1lTVmomJiYVuR5LeMJaOsO77gA8nOQ94C3Ac8HvAsiRL29HCCuCZNv4Z4FRgT5KlwPHAi0P1A4bX6dUlSWMw7yOJqrq0qlZU1UoGF57vqqpfBe4GLmjDNgC3tentbZ62/K6qqla/sN39dBqwCvgGcB+wqt0tdUzbx/b59itJOnyjHEn0/A5wU5LPAw8A17X6dcCNSaaAvQx+6VNVu5PcAjwC7AcuqaofAiT5NHAHsATYWlW7j0K/kqSOIxISVfV14Ott+kkGdybNHPM94COd9S8HLp+lvgPYcSR6lCQdPt9xLUnqMiQkSV2GhCSpy5CQJHUZEpKkLkNCktRlSEiSugwJSVKXISFJ6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkroMCUlSlyEhSeoyJCRJXYaEJKnLkJAkdRkSkqQuQ0KS1DXvkEhyapK7kzySZHeS32r1E5PsTPJE+3pCqyfJVUmmkjyY5IyhbW1o459IsmGo/p4kD7V1rkqSUZ6sJOnwjHIksR/4F1W1GjgLuCTJamAzcGdVrQLubPMA5wKr2mMTcA0MQgW4DHgvcCZw2YFgaWM+MbTeuhH6lSQdpnmHRFU9W1V/2qb/F/AosBxYD2xrw7YB57fp9cANNbALWJbkFOAcYGdV7a2qfcBOYF1bdlxV7aqqAm4Y2pYkaQyOyDWJJCuBdwP3AidX1bNt0XPAyW16OfD00Gp7Wu1g9T2z1Gfb/6Ykk0kmp6enR3sykqQfGTkkkvx14D8D/7Sqvju8rB0B1Kj7mEtVbamqNVW1ZmJi4mjvTpLeNEYKiSR/jUFAfKmqvtLKz7dTRbSvL7T6M8CpQ6uvaLWD1VfMUpckjckodzcFuA54tKp+d2jRduDAHUobgNuG6he3u5zOAl5up6XuAM5OckK7YH02cEdb9t0kZ7V9XTy0LUnSGCwdYd33Ab8GPJTkm632L4EvALck2Qh8G/hoW7YDOA+YAl4BPg5QVXuTfA64r437bFXtbdOfAq4H3grc3h6SpDGZd0hU1f8Aeu9bWDvL+AIu6WxrK7B1lvokcPp8e5QkjcZ3XEuSugwJSVKXISFJ6jIkJEldhoQkqcuQkCR1GRKSpC5DQpLUZUhIkroMCUlSlyEhSeoyJCRJXYaEJKnLkJAkdRkSkqSuUf7TIb0BrNz81QXZ77e+8KEF2a+kw+ORhCSpy5CQJHUZEpKkLq9JaEEs1LUQ8HqIdDgMCb3peLFeOnSebpIkdXkkMWQhT4Hojc9TbHo9WvRHEknWJXksyVSSzQvdjyS9mSzqI4kkS4CrgX8A7AHuS7K9qh5Z2M6k1xevw2i+FnVIAGcCU1X1JECSm4D1gCEhvQ54iu31b7GHxHLg6aH5PcB7Zw5KsgnY1Gb/d5LHxtDbsJOAvxzzPg+Vvc2Pvc3PouktV7ymtGh6m8Vi6O1vzlZc7CFxSKpqC7BlofafZLKq1izU/g/G3ubH3ubH3uZnMfe22C9cPwOcOjS/otUkSWOw2EPiPmBVktOSHANcCGxf4J4k6U1jUZ9uqqr9ST4N3AEsAbZW1e4Fbms2C3aq6xDY2/zY2/zY2/ws2t5SVQvdgyRpkVrsp5skSQvIkJAkdRkS85DkxCQ7kzzRvp4wy5ifT/InSXYneTDJPzrKPR3040uSHJvk5rb83iQrj2Y/h9nbP0/ySHud7kwy6/3aC9Hb0Lh/mKSSjO02xUPpLclH22u3O8nvL5bekvx0kruTPNC+r+eNqa+tSV5I8nBneZJc1fp+MMkZ4+jrEHv71dbTQ0n+OMnPjau3g6oqH4f5AP41sLlNbwaumGXMO4FVbfqngGeBZUepnyXAnwM/AxwD/E9g9YwxnwL+Q5u+ELh5TK/VofT2y8Db2vRvLKbe2rifAO4BdgFrFktvwCrgAeCENv83FlFvW4DfaNOrgW+NqbdfBM4AHu4sPw+4HQhwFnDvOPo6xN7+3tD38txx9nawh0cS87Me2NamtwHnzxxQVY9X1RNt+i+AF4CJo9TPjz6+pKpeBQ58fEmv51uBtUlylPo5rN6q6u6qeqXN7mLwfphxOJTXDeBzwBXA98bU16H29gng6qraB1BVLyyi3go4rk0fD/zFOBqrqnuAvQcZsh64oQZ2AcuSnLIYequqPz7wvWS8PwcHZUjMz8lV9Wybfg44+WCDk5zJ4C+uPz9K/cz28SXLe2Oqaj/wMvD2o9TP4fY2bCODv/TGYc7e2umIU6tq3B9CdCiv2zuBdyb5oyS7kqxbRL39K+BjSfYAO4DfHE9rczrcf48LZZw/Bwe1qN8nsZCSfA34yVkWfWZ4pqoqSfc+4vZXyo3Ahqr6qyPb5RtLko8Ba4BfWuheAJL8GPC7wK8vcCs9Sxmccno/g78670nyd6vqpYVsqrkIuL6q/l2SXwBuTHK6PwNzS/LLDELi7y90L2BIdFXVB3vLkjyf5JSqeraFwKyH+UmOA74KfKYd2h4th/LxJQfG7EmylMEpgBePYk+H0xtJPsgggH+pqr4/hr4OpbefAE4Hvt7OzP0ksD3Jh6tqcoF7g8FfwfdW1Q+Ap5I8ziA07lsEvW0E1gFU1Z8keQuDD7Eb1ymxnkX9UT9Jfha4Fji3qsbx8zknTzfNz3ZgQ5veANw2c0D7GJE/ZHD+89aj3M+hfHzJcM8XAHdVu0K20L0leTfwH4EPj/G8+py9VdXLVXVSVa2sqpUMzhOPIyDm7K35LwyOIkhyEoPTT08ukt6+A6xtvf0d4C3A9Bh6m8t24OJ2l9NZwMtDp44XVJKfBr4C/FpVPb7Q/fzIQl85fz0+GJzLvxN4AvgacGKrrwGubdMfA34AfHPo8fNHsafzgMcZXPf4TKt9lsEvNRj8kP4BMAV8A/iZMb5ec/X2NeD5oddp+2LpbcbYrzOmu5sO8XULg9NhjwAPARcuot5WA3/E4M6nbwJnj6mvLzO4k/AHDI60NgKfBD459Jpd3fp+aMzfz7l6uxbYN/RzMDmu3g728GM5JEldnm6SJHUZEpKkLkNCktRlSEiSugwJSVKXISFJ6jIkJEld/xd/+Ks8BAuwnwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot histogram of the prediction to find a suitable threshold. From the histogram a 0.1 looks like a good choice.\n",
    "plt.hist(a['out'].data.cpu().numpy().flatten())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the input image, ground truth and the predicted output\n",
    "plt.figure(figsize=(10,10));\n",
    "plt.subplot(131);\n",
    "plt.imshow(img[0,...].transpose(1,2,0));\n",
    "plt.title('Image')\n",
    "plt.axis('off');\n",
    "plt.subplot(132);\n",
    "plt.imshow(mask);\n",
    "plt.title('Ground Truth')\n",
    "plt.axis('off');\n",
    "plt.subplot(133);\n",
    "plt.imshow(a['out'].cpu().detach().numpy()[0][0]>0.2);\n",
    "plt.title('Segmentation Output')\n",
    "plt.axis('off');\n",
    "plt.savefig('./CFExp/SegmentationOutput.png',bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "PyTorch",
   "language": "python",
   "name": "torch"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}