ResNet_phi.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "ResNet_phi.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "a0ea46df675c4693a341702eb247afcc": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "state": {
            "_view_name": "HBoxView",
            "_dom_classes": [],
            "_model_name": "HBoxModel",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "box_style": "",
            "layout": "IPY_MODEL_53856ab423944b37829c66c8ea76fadc",
            "_model_module": "@jupyter-widgets/controls",
            "children": [
              "IPY_MODEL_0b1088d38669453282627cc90f07b583",
              "IPY_MODEL_b1638fede3c74fc38def7d1a9d34e204"
            ]
          }
        },
        "53856ab423944b37829c66c8ea76fadc": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "0b1088d38669453282627cc90f07b583": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "IntProgressModel",
          "state": {
            "_view_name": "ProgressView",
            "style": "IPY_MODEL_3fe123cbe1fe47869772197a3ae39a7b",
            "_dom_classes": [],
            "description": "100%",
            "_model_name": "IntProgressModel",
            "bar_style": "success",
            "max": 178728960,
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": 178728960,
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "orientation": "horizontal",
            "min": 0,
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_d0231d5e4ce54b1ea36ae8237d507c47"
          }
        },
        "b1638fede3c74fc38def7d1a9d34e204": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "state": {
            "_view_name": "HTMLView",
            "style": "IPY_MODEL_672f0980babd47ea9dcbf23444616534",
            "_dom_classes": [],
            "description": "",
            "_model_name": "HTMLModel",
            "placeholder": "​",
            "_view_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "value": " 170M/170M [01:50&lt;00:00, 1.62MB/s]",
            "_view_count": null,
            "_view_module_version": "1.5.0",
            "description_tooltip": null,
            "_model_module": "@jupyter-widgets/controls",
            "layout": "IPY_MODEL_6faf65469c7c4fc981dc18ded9cd76cd"
          }
        },
        "3fe123cbe1fe47869772197a3ae39a7b": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "ProgressStyleModel",
            "description_width": "initial",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "bar_color": null,
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "d0231d5e4ce54b1ea36ae8237d507c47": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        },
        "672f0980babd47ea9dcbf23444616534": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_view_name": "StyleView",
            "_model_name": "DescriptionStyleModel",
            "description_width": "",
            "_view_module": "@jupyter-widgets/base",
            "_model_module_version": "1.5.0",
            "_view_count": null,
            "_view_module_version": "1.2.0",
            "_model_module": "@jupyter-widgets/controls"
          }
        },
        "6faf65469c7c4fc981dc18ded9cd76cd": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "state": {
            "_view_name": "LayoutView",
            "grid_template_rows": null,
            "right": null,
            "justify_content": null,
            "_view_module": "@jupyter-widgets/base",
            "overflow": null,
            "_model_module_version": "1.2.0",
            "_view_count": null,
            "flex_flow": null,
            "width": null,
            "min_width": null,
            "border": null,
            "align_items": null,
            "bottom": null,
            "_model_module": "@jupyter-widgets/base",
            "top": null,
            "grid_column": null,
            "overflow_y": null,
            "overflow_x": null,
            "grid_auto_flow": null,
            "grid_area": null,
            "grid_template_columns": null,
            "flex": null,
            "_model_name": "LayoutModel",
            "justify_items": null,
            "grid_row": null,
            "max_height": null,
            "align_content": null,
            "visibility": null,
            "align_self": null,
            "height": null,
            "min_height": null,
            "padding": null,
            "grid_auto_rows": null,
            "grid_gap": null,
            "max_width": null,
            "order": null,
            "_view_module_version": "1.2.0",
            "grid_template_areas": null,
            "object_position": null,
            "object_fit": null,
            "grid_auto_columns": null,
            "margin": null,
            "display": null,
            "left": null
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "h7E8F9ro4I_y",
        "colab_type": "text"
      },
      "source": [
        "**Input:** Initialization Level Set \n",
        "\n",
        "**Loss Function:** MSE "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GqzAQK5pDXmT",
        "colab_type": "text"
      },
      "source": [
        "# Google Drive Mounting"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2R3rpNTeDVyz",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from google.colab import drive"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v186vKGWDjfr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "drive.mount('/content/drive')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cr6rTkcfDwSH",
        "colab_type": "text"
      },
      "source": [
        "# Data Generator"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ADQ-bL-3DkR2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np \n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mnUgZ1u9Gj5n",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import torch\n",
        "from torch.utils.data import Dataset, DataLoader\n",
        "from torchvision import transforms"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E7JGp-EuD6bN",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import scipy.io         # for loading .mat ground truth files \n",
        "from PIL import Image   # for loading images to be resized "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YgvMGI2y9z6-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "eps = np.finfo(float).eps   # a small value, to avoid zero division"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RtcXu7uJFkBH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import scipy.ndimage as nd\n",
        "\n",
        "def bwdist(mask):\n",
        "    return nd.distance_transform_edt(np.logical_not(mask))\n",
        "\n",
        "def mask2levelSet(mask):\n",
        "    return bwdist(mask) - bwdist(1 - mask)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y2AhPcGL-Rgd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#!python generateMasks.py \"/content/drive/My Drive/DeepLevelSet/train_images\" \"/content/drive/My Drive/DeepLevelSet/train_masks\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uhkKG7gRGSmR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Dataset(Dataset):\n",
        "  def __init__(self, list_IDs):\n",
        "        'Initialization'\n",
        "        self.list_IDs = list_IDs\n",
        "        self.base_url = \"/content/drive/My Drive/DeepLevelSet\"\n",
        "\n",
        "  def __len__(self):\n",
        "        'Denotes the total number of samples'\n",
        "        return len(self.list_IDs)\n",
        "\n",
        "  def __getitem__(self, index):\n",
        "        'Generates one sample of data'\n",
        "        ID = self.list_IDs[index]\n",
        "        fileName = ID.split('.')[0]                                 # name without file extension  \n",
        "\n",
        "        X_img = Image.open(f\"{self.base_url}/train_images/{ID}\")    # loading image\n",
        "        X_img = X_img.resize((224, 224))                            # resizing to 224x224 \n",
        "        X_img = transforms.functional.to_tensor(X_img)              # PIL image to Tensor \n",
        "        \n",
        "        X_mask = np.load(f\"{self.base_url}/train_masks/{fileName}.npy\")\n",
        "        X_mask = np.where(X_mask==1, 0, 1)                          # inverting 1 and 0\n",
        "        phi_init = mask2levelSet(X_mask).astype(\"float32\")          # mask to level set\n",
        "        X_mask = transforms.functional.to_tensor(phi_init)          # converting to tensor \n",
        "\n",
        "        X = torch.cat((X_img, X_mask), dim=0)                       # X = [X_img, X_mask]\n",
        "        \n",
        "        yMat = scipy.io.loadmat(f\"{self.base_url}/groundTruth/{fileName}.mat\")\n",
        "        y = yMat['groundTruth'][0][0][0][0][0] - 1                  # fetching ground truth mask \n",
        "        y = Image.fromarray(y)                                      # converting to PIL image \n",
        "        y = y.resize((224, 224))                                    # resizing to 224x224\n",
        "        y = np.asarray(y)                                           # converting back to array \n",
        "        phi_gt =  mask2levelSet(y).astype(\"float32\")                # mask to level set\n",
        "\n",
        "        Y = phi_init - phi_gt                                       # difference of levelsets \n",
        "        Y = Y / (np.abs(Y) + eps)                                   # normalizing \n",
        "        Y = transforms.functional.to_tensor(Y)                      # to tensor \n",
        "        Y = torch.squeeze(Y.view(-1, 1))                            # reshaping into a vector \n",
        "\n",
        "        return X, Y"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Io4k0lFaMGUA",
        "colab_type": "text"
      },
      "source": [
        "# Model Definition"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5rwO9PCYMCOs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import torchvision.models as models"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Fr12ZpdqMJfC",
        "colab_type": "code",
        "outputId": "9b2bf865-0705-49d6-c410-7176fba3178a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 103,
          "referenced_widgets": [
            "a0ea46df675c4693a341702eb247afcc",
            "53856ab423944b37829c66c8ea76fadc",
            "0b1088d38669453282627cc90f07b583",
            "b1638fede3c74fc38def7d1a9d34e204",
            "3fe123cbe1fe47869772197a3ae39a7b",
            "d0231d5e4ce54b1ea36ae8237d507c47",
            "672f0980babd47ea9dcbf23444616534",
            "6faf65469c7c4fc981dc18ded9cd76cd"
          ]
        }
      },
      "source": [
        "# loading pretrained resnet101\n",
        "model = models.resnet101(pretrained=True)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading: \"https://download.pytorch.org/models/resnet101-5d3b4d8f.pth\" to /root/.cache/torch/checkpoints/resnet101-5d3b4d8f.pth\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "a0ea46df675c4693a341702eb247afcc",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, max=178728960), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "c52iE613MNDa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# modifying first conv layer for 4 channel input\n",
        "weight_conv1 = model.conv1.weight.clone()\n",
        "\n",
        "model.conv1 = torch.nn.Conv2d(4, 64, kernel_size=7, stride=2, padding=3, bias=False)\n",
        "with torch.no_grad():\n",
        "    model.conv1.weight[:, :3] = weight_conv1\n",
        "    model.conv1.weight[:, 3] = model.conv1.weight[:, 0]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7mJqKyojMRH1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# modifying last layer \n",
        "model.fc = torch.nn.Sequential(\n",
        "    torch.nn.Linear(in_features=2048, out_features=224*224),\n",
        "    torch.nn.Sigmoid()\n",
        ")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JfkThNz7M_Mu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# unfreezing model layers \n",
        "for name, param in model.named_parameters():\n",
        "    try:\n",
        "        param.requires_grad = True\n",
        "    except Exception as e: \n",
        "        print(f\"{name}: {e}\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GDHVsOIsaOuc",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# loading saved weights \n",
        "# BASE_URL = \"/content/drive/My Drive/DeepLevelSet\"\n",
        "# model.load_state_dict(torch.load(f\"{BASE_URL}/Resnet_phi.pth\"))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qDb8taPNNG3y",
        "colab_type": "code",
        "outputId": "c5f1afd0-d68a-4e8b-85b2-e9d8993b2cce",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "# Use GPU if available\n",
        "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "# moving model to GPU if available\n",
        "model.to(device)"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "ResNet(\n",
              "  (conv1): Conv2d(4, 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=(2, 2), 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=(2, 2), 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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(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",
              "    )\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=(2, 2), padding=(1, 1), 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=(2, 2), 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=(1, 1), 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=(1, 1), 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",
              "  (avgpool): AdaptiveAvgPool2d(output_size=(1, 1))\n",
              "  (fc): Sequential(\n",
              "    (0): Linear(in_features=2048, out_features=50176, bias=True)\n",
              "    (1): Sigmoid()\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gkHkypdyMepP",
        "colab_type": "text"
      },
      "source": [
        "# Data Loader"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qnjGtfiZQl4_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hj8E4WWiMXzs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "dirs = os.listdir(\"/content/drive/My Drive/DeepLevelSet/train_images\")\n",
        "partition = {\n",
        "    'train': dirs\n",
        "    }"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RiRbkxcLMm6a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "params = {\n",
        "    'batch_size': 32,\n",
        "    'shuffle': True,\n",
        "    'num_workers': 6\n",
        "    }\n",
        "\n",
        "training_set = Dataset(partition['train'])\n",
        "training_generator = DataLoader(training_set, **params)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wu7dQejnM5w1",
        "colab_type": "text"
      },
      "source": [
        "# Model Training "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uXk-LdUUNXOX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from torch import optim"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XftfXvZ_M5Gy",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "criterion = torch.nn.MSELoss()\n",
        "optimizer = torch.optim.Adam(params=model.parameters()) "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-yGSW_vbMvuG",
        "colab_type": "code",
        "outputId": "162a5d05-be6f-4584-ae20-d505cb553060",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "epochs = 300\n",
        "\n",
        "train_losses = []\n",
        "for e in range(epochs):\n",
        "    running_loss = 0\n",
        "    for local_batch, local_labels in training_generator:\n",
        "        inputs, labels = local_batch.to(device), local_labels.to(device)\n",
        "        F = model(inputs)\n",
        "        loss = criterion(2*F-1, labels)\n",
        "        optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "        running_loss += loss.item()\n",
        "\n",
        "    train_losses.append(running_loss/len(training_generator))\n",
        "\n",
        "    print(f\"Epoch: {e+1}/{epochs}.. Training Loss: {running_loss/len(training_generator):.4f}\")"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch: 1/300.. Training Loss: 0.8042\n",
            "Epoch: 2/300.. Training Loss: 0.7247\n",
            "Epoch: 3/300.. Training Loss: 0.6852\n",
            "Epoch: 4/300.. Training Loss: 0.6400\n",
            "Epoch: 5/300.. Training Loss: 0.6161\n",
            "Epoch: 6/300.. Training Loss: 0.6011\n",
            "Epoch: 7/300.. Training Loss: 0.5839\n",
            "Epoch: 8/300.. Training Loss: 0.5946\n",
            "Epoch: 9/300.. Training Loss: 0.5794\n",
            "Epoch: 10/300.. Training Loss: 0.5726\n",
            "Epoch: 11/300.. Training Loss: 0.5572\n",
            "Epoch: 12/300.. Training Loss: 0.5454\n",
            "Epoch: 13/300.. Training Loss: 0.5476\n",
            "Epoch: 14/300.. Training Loss: 0.5453\n",
            "Epoch: 15/300.. Training Loss: 0.5631\n",
            "Epoch: 16/300.. Training Loss: 0.5484\n",
            "Epoch: 17/300.. Training Loss: 0.5586\n",
            "Epoch: 18/300.. Training Loss: 0.5245\n",
            "Epoch: 19/300.. Training Loss: 0.5105\n",
            "Epoch: 20/300.. Training Loss: 0.5067\n",
            "Epoch: 21/300.. Training Loss: 0.4898\n",
            "Epoch: 22/300.. Training Loss: 0.4692\n",
            "Epoch: 23/300.. Training Loss: 0.4578\n",
            "Epoch: 24/300.. Training Loss: 0.4475\n",
            "Epoch: 25/300.. Training Loss: 0.4456\n",
            "Epoch: 26/300.. Training Loss: 0.4268\n",
            "Epoch: 27/300.. Training Loss: 0.4432\n",
            "Epoch: 28/300.. Training Loss: 0.4184\n",
            "Epoch: 29/300.. Training Loss: 0.3927\n",
            "Epoch: 30/300.. Training Loss: 0.3876\n",
            "Epoch: 31/300.. Training Loss: 0.3780\n",
            "Epoch: 32/300.. Training Loss: 0.3573\n",
            "Epoch: 33/300.. Training Loss: 0.3265\n",
            "Epoch: 34/300.. Training Loss: 0.3251\n",
            "Epoch: 35/300.. Training Loss: 0.3303\n",
            "Epoch: 36/300.. Training Loss: 0.3212\n",
            "Epoch: 37/300.. Training Loss: 0.3068\n",
            "Epoch: 38/300.. Training Loss: 0.3041\n",
            "Epoch: 39/300.. Training Loss: 0.2896\n",
            "Epoch: 40/300.. Training Loss: 0.2850\n",
            "Epoch: 41/300.. Training Loss: 0.2839\n",
            "Epoch: 42/300.. Training Loss: 0.2781\n",
            "Epoch: 43/300.. Training Loss: 0.2517\n",
            "Epoch: 44/300.. Training Loss: 0.2354\n",
            "Epoch: 45/300.. Training Loss: 0.2385\n",
            "Epoch: 46/300.. Training Loss: 0.2366\n",
            "Epoch: 47/300.. Training Loss: 0.2387\n",
            "Epoch: 48/300.. Training Loss: 0.2281\n",
            "Epoch: 49/300.. Training Loss: 0.2138\n",
            "Epoch: 50/300.. Training Loss: 0.2062\n",
            "Epoch: 51/300.. Training Loss: 0.2176\n",
            "Epoch: 52/300.. Training Loss: 0.2159\n",
            "Epoch: 53/300.. Training Loss: 0.2107\n",
            "Epoch: 54/300.. Training Loss: 0.1986\n",
            "Epoch: 55/300.. Training Loss: 0.1991\n",
            "Epoch: 56/300.. Training Loss: 0.1847\n",
            "Epoch: 57/300.. Training Loss: 0.1948\n",
            "Epoch: 58/300.. Training Loss: 0.1845\n",
            "Epoch: 59/300.. Training Loss: 0.1830\n",
            "Epoch: 60/300.. Training Loss: 0.1782\n",
            "Epoch: 61/300.. Training Loss: 0.1630\n",
            "Epoch: 62/300.. Training Loss: 0.1642\n",
            "Epoch: 63/300.. Training Loss: 0.1823\n",
            "Epoch: 64/300.. Training Loss: 0.1657\n",
            "Epoch: 65/300.. Training Loss: 0.1554\n",
            "Epoch: 66/300.. Training Loss: 0.1629\n",
            "Epoch: 67/300.. Training Loss: 0.1646\n",
            "Epoch: 68/300.. Training Loss: 0.1499\n",
            "Epoch: 69/300.. Training Loss: 0.1537\n",
            "Epoch: 70/300.. Training Loss: 0.1490\n",
            "Epoch: 71/300.. Training Loss: 0.1464\n",
            "Epoch: 72/300.. Training Loss: 0.1480\n",
            "Epoch: 73/300.. Training Loss: 0.1322\n",
            "Epoch: 74/300.. Training Loss: 0.1403\n",
            "Epoch: 75/300.. Training Loss: 0.1349\n",
            "Epoch: 76/300.. Training Loss: 0.1400\n",
            "Epoch: 77/300.. Training Loss: 0.1484\n",
            "Epoch: 78/300.. Training Loss: 0.1507\n",
            "Epoch: 79/300.. Training Loss: 0.1652\n",
            "Epoch: 80/300.. Training Loss: 0.1437\n",
            "Epoch: 81/300.. Training Loss: 0.1405\n",
            "Epoch: 82/300.. Training Loss: 0.1431\n",
            "Epoch: 83/300.. Training Loss: 0.1300\n",
            "Epoch: 84/300.. Training Loss: 0.1287\n",
            "Epoch: 85/300.. Training Loss: 0.1156\n",
            "Epoch: 86/300.. Training Loss: 0.1193\n",
            "Epoch: 87/300.. Training Loss: 0.1173\n",
            "Epoch: 88/300.. Training Loss: 0.1145\n",
            "Epoch: 89/300.. Training Loss: 0.1201\n",
            "Epoch: 90/300.. Training Loss: 0.1104\n",
            "Epoch: 91/300.. Training Loss: 0.1076\n",
            "Epoch: 92/300.. Training Loss: 0.1061\n",
            "Epoch: 93/300.. Training Loss: 0.1053\n",
            "Epoch: 94/300.. Training Loss: 0.1120\n",
            "Epoch: 95/300.. Training Loss: 0.1064\n",
            "Epoch: 96/300.. Training Loss: 0.1154\n",
            "Epoch: 97/300.. Training Loss: 0.1076\n",
            "Epoch: 98/300.. Training Loss: 0.0923\n",
            "Epoch: 99/300.. Training Loss: 0.0947\n",
            "Epoch: 100/300.. Training Loss: 0.1030\n",
            "Epoch: 101/300.. Training Loss: 0.1097\n",
            "Epoch: 102/300.. Training Loss: 0.1078\n",
            "Epoch: 103/300.. Training Loss: 0.0996\n",
            "Epoch: 104/300.. Training Loss: 0.0906\n",
            "Epoch: 105/300.. Training Loss: 0.0937\n",
            "Epoch: 106/300.. Training Loss: 0.1074\n",
            "Epoch: 107/300.. Training Loss: 0.0892\n",
            "Epoch: 108/300.. Training Loss: 0.0931\n",
            "Epoch: 109/300.. Training Loss: 0.0849\n",
            "Epoch: 110/300.. Training Loss: 0.0768\n",
            "Epoch: 111/300.. Training Loss: 0.0874\n",
            "Epoch: 112/300.. Training Loss: 0.0856\n",
            "Epoch: 113/300.. Training Loss: 0.0758\n",
            "Epoch: 114/300.. Training Loss: 0.0905\n",
            "Epoch: 115/300.. Training Loss: 0.0854\n",
            "Epoch: 116/300.. Training Loss: 0.0769\n",
            "Epoch: 117/300.. Training Loss: 0.0859\n",
            "Epoch: 118/300.. Training Loss: 0.0745\n",
            "Epoch: 119/300.. Training Loss: 0.0701\n",
            "Epoch: 120/300.. Training Loss: 0.0814\n",
            "Epoch: 121/300.. Training Loss: 0.0663\n",
            "Epoch: 122/300.. Training Loss: 0.0805\n",
            "Epoch: 123/300.. Training Loss: 0.0720\n",
            "Epoch: 124/300.. Training Loss: 0.0658\n",
            "Epoch: 125/300.. Training Loss: 0.0669\n",
            "Epoch: 126/300.. Training Loss: 0.0676\n",
            "Epoch: 127/300.. Training Loss: 0.0681\n",
            "Epoch: 128/300.. Training Loss: 0.0731\n",
            "Epoch: 129/300.. Training Loss: 0.0642\n",
            "Epoch: 130/300.. Training Loss: 0.0708\n",
            "Epoch: 131/300.. Training Loss: 0.0597\n",
            "Epoch: 132/300.. Training Loss: 0.0651\n",
            "Epoch: 133/300.. Training Loss: 0.0566\n",
            "Epoch: 134/300.. Training Loss: 0.0644\n",
            "Epoch: 135/300.. Training Loss: 0.0629\n",
            "Epoch: 136/300.. Training Loss: 0.0671\n",
            "Epoch: 137/300.. Training Loss: 0.0557\n",
            "Epoch: 138/300.. Training Loss: 0.0539\n",
            "Epoch: 139/300.. Training Loss: 0.0563\n",
            "Epoch: 140/300.. Training Loss: 0.0538\n",
            "Epoch: 141/300.. Training Loss: 0.0557\n",
            "Epoch: 142/300.. Training Loss: 0.0514\n",
            "Epoch: 143/300.. Training Loss: 0.0503\n",
            "Epoch: 144/300.. Training Loss: 0.0448\n",
            "Epoch: 145/300.. Training Loss: 0.0477\n",
            "Epoch: 146/300.. Training Loss: 0.0456\n",
            "Epoch: 147/300.. Training Loss: 0.0425\n",
            "Epoch: 148/300.. Training Loss: 0.0578\n",
            "Epoch: 149/300.. Training Loss: 0.0540\n",
            "Epoch: 150/300.. Training Loss: 0.0488\n",
            "Epoch: 151/300.. Training Loss: 0.0467\n",
            "Epoch: 152/300.. Training Loss: 0.0462\n",
            "Epoch: 153/300.. Training Loss: 0.0461\n",
            "Epoch: 154/300.. Training Loss: 0.0448\n",
            "Epoch: 155/300.. Training Loss: 0.0430\n",
            "Epoch: 156/300.. Training Loss: 0.0425\n",
            "Epoch: 157/300.. Training Loss: 0.0467\n",
            "Epoch: 158/300.. Training Loss: 0.0435\n",
            "Epoch: 159/300.. Training Loss: 0.0425\n",
            "Epoch: 160/300.. Training Loss: 0.0398\n",
            "Epoch: 161/300.. Training Loss: 0.0444\n",
            "Epoch: 162/300.. Training Loss: 0.0345\n",
            "Epoch: 163/300.. Training Loss: 0.0374\n",
            "Epoch: 164/300.. Training Loss: 0.0348\n",
            "Epoch: 165/300.. Training Loss: 0.0367\n",
            "Epoch: 166/300.. Training Loss: 0.0433\n",
            "Epoch: 167/300.. Training Loss: 0.0512\n",
            "Epoch: 168/300.. Training Loss: 0.0448\n",
            "Epoch: 169/300.. Training Loss: 0.0443\n",
            "Epoch: 170/300.. Training Loss: 0.0409\n",
            "Epoch: 171/300.. Training Loss: 0.0445\n",
            "Epoch: 172/300.. Training Loss: 0.0374\n",
            "Epoch: 173/300.. Training Loss: 0.0360\n",
            "Epoch: 174/300.. Training Loss: 0.0308\n",
            "Epoch: 175/300.. Training Loss: 0.0299\n",
            "Epoch: 176/300.. Training Loss: 0.0321\n",
            "Epoch: 177/300.. Training Loss: 0.0334\n",
            "Epoch: 178/300.. Training Loss: 0.0338\n",
            "Epoch: 179/300.. Training Loss: 0.0325\n",
            "Epoch: 180/300.. Training Loss: 0.0307\n",
            "Epoch: 181/300.. Training Loss: 0.0312\n",
            "Epoch: 182/300.. Training Loss: 0.0329\n",
            "Epoch: 183/300.. Training Loss: 0.0299\n",
            "Epoch: 184/300.. Training Loss: 0.0270\n",
            "Epoch: 185/300.. Training Loss: 0.0271\n",
            "Epoch: 186/300.. Training Loss: 0.0247\n",
            "Epoch: 187/300.. Training Loss: 0.0256\n",
            "Epoch: 188/300.. Training Loss: 0.0222\n",
            "Epoch: 189/300.. Training Loss: 0.0216\n",
            "Epoch: 190/300.. Training Loss: 0.0229\n",
            "Epoch: 191/300.. Training Loss: 0.0216\n",
            "Epoch: 192/300.. Training Loss: 0.0225\n",
            "Epoch: 193/300.. Training Loss: 0.0239\n",
            "Epoch: 194/300.. Training Loss: 0.0258\n",
            "Epoch: 195/300.. Training Loss: 0.0247\n",
            "Epoch: 196/300.. Training Loss: 0.0236\n",
            "Epoch: 197/300.. Training Loss: 0.0242\n",
            "Epoch: 198/300.. Training Loss: 0.0223\n",
            "Epoch: 199/300.. Training Loss: 0.0187\n",
            "Epoch: 200/300.. Training Loss: 0.0215\n",
            "Epoch: 201/300.. Training Loss: 0.0159\n",
            "Epoch: 202/300.. Training Loss: 0.0181\n",
            "Epoch: 203/300.. Training Loss: 0.0192\n",
            "Epoch: 204/300.. Training Loss: 0.0195\n",
            "Epoch: 205/300.. Training Loss: 0.0173\n",
            "Epoch: 206/300.. Training Loss: 0.0182\n",
            "Epoch: 207/300.. Training Loss: 0.0189\n",
            "Epoch: 208/300.. Training Loss: 0.0165\n",
            "Epoch: 209/300.. Training Loss: 0.0168\n",
            "Epoch: 210/300.. Training Loss: 0.0196\n",
            "Epoch: 211/300.. Training Loss: 0.0165\n",
            "Epoch: 212/300.. Training Loss: 0.0172\n",
            "Epoch: 213/300.. Training Loss: 0.0172\n",
            "Epoch: 214/300.. Training Loss: 0.0140\n",
            "Epoch: 215/300.. Training Loss: 0.0150\n",
            "Epoch: 216/300.. Training Loss: 0.0178\n",
            "Epoch: 217/300.. Training Loss: 0.0181\n",
            "Epoch: 218/300.. Training Loss: 0.0139\n",
            "Epoch: 219/300.. Training Loss: 0.0170\n",
            "Epoch: 220/300.. Training Loss: 0.0142\n",
            "Epoch: 221/300.. Training Loss: 0.0141\n",
            "Epoch: 222/300.. Training Loss: 0.0201\n",
            "Epoch: 223/300.. Training Loss: 0.0161\n",
            "Epoch: 224/300.. Training Loss: 0.0147\n",
            "Epoch: 225/300.. Training Loss: 0.0177\n",
            "Epoch: 226/300.. Training Loss: 0.0168\n",
            "Epoch: 227/300.. Training Loss: 0.0164\n",
            "Epoch: 228/300.. Training Loss: 0.0139\n",
            "Epoch: 229/300.. Training Loss: 0.0142\n",
            "Epoch: 230/300.. Training Loss: 0.0126\n",
            "Epoch: 231/300.. Training Loss: 0.0117\n",
            "Epoch: 232/300.. Training Loss: 0.0145\n",
            "Epoch: 233/300.. Training Loss: 0.0103\n",
            "Epoch: 234/300.. Training Loss: 0.0121\n",
            "Epoch: 235/300.. Training Loss: 0.0109\n",
            "Epoch: 236/300.. Training Loss: 0.0101\n",
            "Epoch: 237/300.. Training Loss: 0.0102\n",
            "Epoch: 238/300.. Training Loss: 0.0087\n",
            "Epoch: 239/300.. Training Loss: 0.0125\n",
            "Epoch: 240/300.. Training Loss: 0.0126\n",
            "Epoch: 241/300.. Training Loss: 0.0119\n",
            "Epoch: 242/300.. Training Loss: 0.0115\n",
            "Epoch: 243/300.. Training Loss: 0.0119\n",
            "Epoch: 244/300.. Training Loss: 0.0134\n",
            "Epoch: 245/300.. Training Loss: 0.0095\n",
            "Epoch: 246/300.. Training Loss: 0.0103\n",
            "Epoch: 247/300.. Training Loss: 0.0096\n",
            "Epoch: 248/300.. Training Loss: 0.0093\n",
            "Epoch: 249/300.. Training Loss: 0.0112\n",
            "Epoch: 250/300.. Training Loss: 0.0103\n",
            "Epoch: 251/300.. Training Loss: 0.0091\n",
            "Epoch: 252/300.. Training Loss: 0.0081\n",
            "Epoch: 253/300.. Training Loss: 0.0078\n",
            "Epoch: 254/300.. Training Loss: 0.0089\n",
            "Epoch: 255/300.. Training Loss: 0.0077\n",
            "Epoch: 256/300.. Training Loss: 0.0079\n",
            "Epoch: 257/300.. Training Loss: 0.0090\n",
            "Epoch: 258/300.. Training Loss: 0.0078\n",
            "Epoch: 259/300.. Training Loss: 0.0076\n",
            "Epoch: 260/300.. Training Loss: 0.0091\n",
            "Epoch: 261/300.. Training Loss: 0.0072\n",
            "Epoch: 262/300.. Training Loss: 0.0068\n",
            "Epoch: 263/300.. Training Loss: 0.0062\n",
            "Epoch: 264/300.. Training Loss: 0.0093\n",
            "Epoch: 265/300.. Training Loss: 0.0077\n",
            "Epoch: 266/300.. Training Loss: 0.0076\n",
            "Epoch: 267/300.. Training Loss: 0.0070\n",
            "Epoch: 268/300.. Training Loss: 0.0064\n",
            "Epoch: 269/300.. Training Loss: 0.0063\n",
            "Epoch: 270/300.. Training Loss: 0.0047\n",
            "Epoch: 271/300.. Training Loss: 0.0055\n",
            "Epoch: 272/300.. Training Loss: 0.0052\n",
            "Epoch: 273/300.. Training Loss: 0.0065\n",
            "Epoch: 274/300.. Training Loss: 0.0063\n",
            "Epoch: 275/300.. Training Loss: 0.0061\n",
            "Epoch: 276/300.. Training Loss: 0.0070\n",
            "Epoch: 277/300.. Training Loss: 0.0070\n",
            "Epoch: 278/300.. Training Loss: 0.0072\n",
            "Epoch: 279/300.. Training Loss: 0.0058\n",
            "Epoch: 280/300.. Training Loss: 0.0057\n",
            "Epoch: 281/300.. Training Loss: 0.0046\n",
            "Epoch: 282/300.. Training Loss: 0.0051\n",
            "Epoch: 283/300.. Training Loss: 0.0051\n",
            "Epoch: 284/300.. Training Loss: 0.0061\n",
            "Epoch: 285/300.. Training Loss: 0.0055\n",
            "Epoch: 286/300.. Training Loss: 0.0051\n",
            "Epoch: 287/300.. Training Loss: 0.0051\n",
            "Epoch: 288/300.. Training Loss: 0.0056\n",
            "Epoch: 289/300.. Training Loss: 0.0048\n",
            "Epoch: 290/300.. Training Loss: 0.0043\n",
            "Epoch: 291/300.. Training Loss: 0.0047\n",
            "Epoch: 292/300.. Training Loss: 0.0040\n",
            "Epoch: 293/300.. Training Loss: 0.0044\n",
            "Epoch: 294/300.. Training Loss: 0.0045\n",
            "Epoch: 295/300.. Training Loss: 0.0051\n",
            "Epoch: 296/300.. Training Loss: 0.0044\n",
            "Epoch: 297/300.. Training Loss: 0.0048\n",
            "Epoch: 298/300.. Training Loss: 0.0041\n",
            "Epoch: 299/300.. Training Loss: 0.0042\n",
            "Epoch: 300/300.. Training Loss: 0.0035\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RDvZ4W24NzVf",
        "colab_type": "code",
        "outputId": "e7120c26-323c-4afb-db6d-d5a9354ce9c3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        }
      },
      "source": [
        "plt.plot(train_losses)\n",
        "plt.title('Model loss')\n",
        "plt.ylabel('Loss')\n",
        "plt.xlabel('Epoch')\n",
        "plt.legend(['Train Loss'])\n",
        "plt.show()"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "561-hqsu_d9O",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# saving model \n",
        "BASE_URL = \"/content/drive/My Drive/DeepLevelSet\"\n",
        "\n",
        "torch.save(model.state_dict(), f\"{BASE_URL}/ResNet_phi.pth\" )"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tsirq8dgbOnm",
        "colab_type": "text"
      },
      "source": [
        "# Visualizing Output"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MVlewS7F36r9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "BASE_URL = \"/content/drive/My Drive/DeepLevelSet\""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xNDRLmJT2Z7V",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def test_preprocess(img, phi): \n",
        "    X_img = transforms.functional.to_tensor(img)  # PIL image to Tensor\n",
        "    X_mask = transforms.functional.to_tensor(phi.astype(\"float32\") )    # converting to tensor \n",
        "    X = torch.cat((X_img, X_mask), dim=0)   # X = [X_img, X_mask]\n",
        "\n",
        "    return X"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XGe34ud53Id7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_target(ID, phi_init):\n",
        "    yMat = scipy.io.loadmat(f\"{BASE_URL}/groundTruth/{ID}.mat\")\n",
        "    y = yMat['groundTruth'][0][0][0][0][0] - 1                  # fetching ground truth mask \n",
        "    y = Image.fromarray(y)                                      # converting to PIL image \n",
        "    y = y.resize((224, 224))                                    # resizing to 224x224\n",
        "    y = np.asarray(y)                                           # converting back to array \n",
        "    phi_gt =  mask2levelSet(y).astype(\"float32\")                # mask to level set\n",
        "\n",
        "    Y = phi_init - phi_gt                                       # difference of levelsets \n",
        "    Y = Y / (np.abs(Y) + eps)                                   # normalizing \n",
        "\n",
        "    return Y "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YjNLgB2i2aKT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_prediction(X_img, phi): \n",
        "    with torch.no_grad():\n",
        "        model.eval()\n",
        "\n",
        "        # preperaing model input \n",
        "        X = test_preprocess(X_img, phi)\n",
        "        X = X.unsqueeze(0)\n",
        "        X = X.to(device)\n",
        "        \n",
        "        # getting model prediction \n",
        "        F = 2*model(X)-1\n",
        "        F = F.view(224, 224)\n",
        "        F = F.cpu().numpy()\n",
        "\n",
        "        return F"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rAq7Cy3t2aUq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def plot_pred(ID):\n",
        "    # reading image \n",
        "    X_img = Image.open(f\"{BASE_URL}/train_images/{ID}.jpg\")     # loading image\n",
        "    X_img = X_img.resize((224, 224))                            # resizing to 224x224\n",
        "\n",
        "    # reading initialization mask\n",
        "    mask_init = np.load(f\"{BASE_URL}/train_masks/{ID}.npy\")\n",
        "    mask_init = np.where(mask_init==1, 0, 1)\n",
        "    phi = mask2levelSet(mask_init)                              # mask to phi \n",
        "\n",
        "    # getting target \n",
        "    Y = get_target(ID, phi)\n",
        "\n",
        "    # getting model prediction \n",
        "    F = get_prediction(X_img, phi)\n",
        "\n",
        "    # plotting results \n",
        "    plt.figure(figsize=(10, 5))\n",
        "    plt.subplot(2, 2, 1).set_title(\"Image\")\n",
        "    plt.imshow(X_img, cmap='gray')\n",
        "\n",
        "    plt.subplot(2, 2, 2).set_title(\"Init LS\")\n",
        "    plt.imshow(phi, cmap='gray')\n",
        "\n",
        "    plt.subplot(2, 2, 3).set_title(\"Model Predicton\")\n",
        "    plt.imshow(F, cmap='gray')\n",
        "\n",
        "    plt.subplot(2, 2, 4).set_title(\"Target\")\n",
        "    plt.imshow(Y, cmap='gray')\n",
        "\n",
        "\n",
        "    # adding padding between two figures \n",
        "    plt.subplots_adjust(wspace = 0.5)\n",
        "    plt.subplots_adjust(hspace = 0.5)\n",
        "\n",
        "    # displaying figure\n",
        "    plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VQ6d5guRc1sv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "0a52985c-20f9-4a9e-b7f7-e329b0568ed1"
      },
      "source": [
        "plot_pred(1)"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_HSNQSAt4FYB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "9a6e4175-811e-4fd9-c660-14a927ebd90b"
      },
      "source": [
        "plot_pred(2)"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "a28OFUYg9W2P",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "28cc3f69-9040-4d46-91ce-d154f3d24256"
      },
      "source": [
        "plot_pred(3)"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_ws3v5Ep9g_l",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 336
        },
        "outputId": "b5af81a7-a57a-4874-faab-14a35afaec37"
      },
      "source": [
        "plot_pred(5)"
      ],
      "execution_count": 44,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 720x360 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DO6nXh0w9im6",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}