{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"## **Contents**\n-  [Introduction](#i)\n-  [1.Importing Libraries](#1)\n-  [2.Helper Functions](#2)\n-  [3.Dataset Managament](#3)\n    -  [3.1.Dataset and DataLoaders](#3.1)\n    -  [3.2.Visualizing Dataset](#3.2)\n-  [4.Initializing pre-trained model](#4)\n-  [5.Training](#5)\n-  [6.Plotting Graphs](#6)\n    -  [6.1.Plotting Loss vs Epoch](#6.1)\n-  [7.Loading and Testing](#7)","metadata":{"execution":{"iopub.execute_input":"2021-10-22T02:20:53.048499Z","iopub.status.busy":"2021-10-22T02:20:53.047964Z","iopub.status.idle":"2021-10-22T02:20:55.442888Z","shell.execute_reply":"2021-10-22T02:20:55.442186Z","shell.execute_reply.started":"2021-10-22T02:20:53.048407Z"}}},{"cell_type":"markdown","source":"## **Introduction** <a class=\"anchor\" id=\"i\"></a>\n\n\nIn this notebook I am visualizing the dataset and also fine-tuning pre-trained [Mask-RCNN](https://arxiv.org/abs/1703.06870) model using [Pytorch](https://pytorch.org/) library.Mask R-CNN is a popular deep learning instance segmentation technique that performs pixel-level segmentation on detected objects.\n\nThe Mask R-CNN algorithm can accommodate multiple classes and overlapping objects.Mask R-CNN extends Faster R-CNN to solve instance segmentation tasks. It achieves this by adding a branch for predicting an object mask in parallel with the existing branch for bounding box recognition. In principle, Mask R-CNN is an intuitive extension of Faster R-CNN, but constructing the mask branch properly is critical for good results. \n\nInstance segmentation treats multiple objects of the same class as distinct individual instances which give this model slight advantage over other Semantic segmentation models. \n\n\nRead more about Mask-RCNN [here](https://www.analyticsvidhya.com/blog/2019/07/computer-vision-implementing-mask-r-cnn-image-segmentation/)\n\nSome of the helper functions and dataloader classes are the derivatives of [this](https://www.kaggle.com/julian3833/sartorius-starter-torch-mask-r-cnn-lb-0-173#%F0%9F%A6%A0-Sartorius---Starter-Torch-Mask-R-CNN) notebook.\n\n<img src= \"attachment:e21e8027-d32b-49eb-8685-18e763bda774.png\"  style='width: 800px;'>","metadata":{},"attachments":{"e21e8027-d32b-49eb-8685-18e763bda774.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"### **1.Importing Libraries** <a class=\"anchor\" id=\"1\"></a>","metadata":{}},{"cell_type":"code","source":"import os\nimport sys\nimport torch\nimport random\nimport numpy as np\nimport torchvision\nimport pandas as pd\nimport seaborn as sns\nfrom PIL import Image\nfrom PIL import Image\nfrom sklearn import cluster\nimport matplotlib.pyplot as plt\nfrom torch.utils.data import Dataset as Dataset\nfrom torch.utils.data import DataLoader as DataLoader\nfrom torchvision.models.detection.faster_rcnn import FastRCNNPredictor\nfrom torchvision.models.detection.mask_rcnn import MaskRCNNPredictor\n\nsys.path.append(\"../input/maskrcnn-utils/\")\nfrom transforms import ToTensor, RandomHorizontalFlip, Compose\n\ntorch.cuda.empty_cache()\nrandom.seed(0)\nnp.random.seed(0)\ntorch.manual_seed(0)\n\nsns.set_style(\"darkgrid\")","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:05.427174Z","iopub.execute_input":"2023-03-30T17:30:05.427864Z","iopub.status.idle":"2023-03-30T17:30:08.216147Z","shell.execute_reply.started":"2023-03-30T17:30:05.427767Z","shell.execute_reply":"2023-03-30T17:30:08.215373Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **2.Helper Functions** <a class=\"anchor\" id=\"2\"></a>","metadata":{}},{"cell_type":"code","source":"IMG_W ,IMG_H = 520, 704\nMOMENTUM = 0.9\nLEARNING_RATE = 0.001\nWEIGHT_DECAY = 0.0005\n\nroot = \"../input/sartorius-cell-instance-segmentation/\"\n\ndef rle_decode(mask_rle, shape, color=1):\n    s = mask_rle.split()\n    starts, lengths = [np.asarray(x, dtype=int) for x in (s[0:][::2], s[1:][::2])]\n    \n    starts -= 1\n    ends = starts + lengths\n    img = np.zeros(shape[0] * shape[1], dtype=np.float32)\n    for lo, hi in zip(starts, ends):\n        img[lo : hi] = color\n    return img.reshape(shape)\n\ndef test_model(model, dataloader,n):\n    \n    for i in range(n):\n        fig = plt.figure(figsize=(10,10))\n        img, targets = next(iter(dataloader))\n        img = img[0]\n        targets = targets[0]\n        plt.imshow(img.numpy().transpose((1,2,0)))\n        plt.grid(None)\n        plt.title(f\"Image : {i}\")\n        plt.show()\n    \n        fig = plt.figure(figsize=(10,10))\n        all_masks = np.zeros((IMG_W ,IMG_H))\n        for mask in targets['masks']:\n            all_masks = np.logical_or(all_masks, mask)\n        plt.imshow(img.numpy().transpose((1,2,0)))\n        plt.imshow(all_masks, alpha=0.3)\n        plt.grid(None)\n        plt.title(f\"Target : {i}\")\n        plt.show()\n    \n        fig = plt.figure(figsize=(10,10))\n        model.eval()\n        with torch.no_grad():\n            preds = model([img.to(device)])[0]\n\n        plt.imshow(img.cpu().numpy().transpose((1,2,0)))\n        all_preds_masks = np.zeros((IMG_W ,IMG_H))\n        for mask in preds['masks'].cpu().detach().numpy():\n            all_preds_masks = np.logical_or(all_preds_masks, mask[0])\n        plt.imshow(all_preds_masks, alpha=0.4)\n        plt.grid(None)\n        plt.title(f\"Predictions : {i}\")\n        plt.show()\n        \n\n\n# Stolen from: https://www.kaggle.com/arunamenon/cell-instance-segmentation-unet-eda\n# Run-length encoding stolen from https://www.kaggle.com/rakhlin/fast-run-length-encoding-python\n# Modified by me\n\ndef rle_encoding(x):\n    dots = np.where(x.flatten() == 1)[0]\n    run_lengths = []\n    prev = -2\n    for b in dots:\n        if (b>prev+1): run_lengths.extend((b + 1, 0))\n        run_lengths[-1] += 1\n        prev = b\n    return ' '.join(map(str, run_lengths))\n\n\ndef does_overlap(mask, other_masks):\n    for other_mask in other_masks:\n        if np.sum(np.logical_and(mask, other_mask)) > 0:\n            return True\n    return False\n\n\ndef remove_overlapping_pixels(mask, other_masks):\n    for other_mask in other_masks:\n        if np.sum(np.logical_and(mask, other_mask)) > 0:\n            mask[np.logical_and(mask, other_mask)] = 0\n    return mask","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:08.218133Z","iopub.execute_input":"2023-03-30T17:30:08.218414Z","iopub.status.idle":"2023-03-30T17:30:08.237622Z","shell.execute_reply.started":"2023-03-30T17:30:08.218379Z","shell.execute_reply":"2023-03-30T17:30:08.236800Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **3.Dataset Managament** <a class=\"anchor\" id=\"3\"></a>","metadata":{}},{"cell_type":"markdown","source":"#### 1.Downloading and Extracting Dataset <a class=\"anchor\" id=\"3.1\"></a>","metadata":{}},{"cell_type":"code","source":"class SatoriusDataset(torch.utils.data.Dataset):\n    def __init__(self, transforms, root = 'data/',train = True):\n        self.root = root\n        self.transforms = transforms\n        self.w , self.h = 520 , 704\n        info = pd.read_csv(self.root+'train.csv')[[\"id\",\"annotation\"]]\n        info = info.groupby('id')['annotation'].agg(lambda x: list(x)).reset_index()\n        validation =  6\n        self.data_info = 0\n\n        if train == True:\n            self.data_info = info[validation:].reset_index(drop=True)\n        else:\n            self.data_info = info[:validation].reset_index(drop=True)\n            \n    def __getitem__(self, idx):\n        img = Image.open(self.root+'train/'+self.data_info['id'][idx]+'.png').convert(\"RGB\")\n        mask = np.zeros((len(self.data_info['annotation'][idx]), self.w, self.h), dtype=int)\n        \n        num_objs = len(self.data_info['annotation'][idx])\n        \n        for i in range(num_objs):\n            nth_mask = rle_decode(self.data_info['annotation'][idx][i], (self.w, self.h))\n            nth_mask = np.array(nth_mask) > 0\n            mask[i, :, :] = nth_mask\n    \n\n        \n        boxes = []\n        new_masks = []\n\n        for i in range(num_objs):\n            pos = np.where(mask[i, :, :])\n            xmin = np.min(pos[1])\n            xmax = np.max(pos[1])\n            ymin = np.min(pos[0])\n            ymax = np.max(pos[0])\n            boxes.append([xmin, ymin, xmax, ymax])\n            new_masks.append(mask[i, :, :])\n        \n        nmx = np.zeros((num_objs ,self.w, self.h), dtype=int)\n        \n        for i in range(num_objs):\n            nmx[i, :, :] = new_masks[i]\n        \n            \n        boxes = torch.as_tensor(boxes, dtype=torch.float32)\n        labels = torch.ones((num_objs,), dtype=torch.int64)\n        masks = torch.as_tensor(nmx, dtype=torch.uint8)\n\n        image_id = torch.tensor([idx])\n        area = (boxes[:, 3] - boxes[:, 1]) * (boxes[:, 2] - boxes[:, 0])\n        iscrowd = torch.zeros((num_objs,), dtype=torch.int64)\n\n        target = {}\n        target[\"boxes\"] = boxes\n        target[\"labels\"] = labels\n        target[\"masks\"] = masks\n        target[\"image_id\"] = image_id\n        target[\"area\"] = area\n        target[\"iscrowd\"] = iscrowd\n\n        if self.transforms is not None:\n            img,target = self.transforms(img,target)\n            \n        return img, target\n\n    def __len__(self):\n        return len(self.data_info)\n\n\n    \ndef get_transform(train):\n    transforms = []\n    # converts the image, a PIL image, into a PyTorch Tensor\n    transforms.append(ToTensor())\n    if train:\n        # during training, randomly flip the training images\n        # and ground-truth for data augmentation\n#         transforms.append(RandomHorizontalFlip(0.5))\n#         transforms.append(RandomVerticalFlip(0.5))\n        pass\n    return Compose(transforms)\n\n\ntrain_dataset = SatoriusDataset(transforms=get_transform(train=True),root =root,train = True)\ntrain_dataloader = DataLoader(train_dataset, batch_size=2, shuffle=True, collate_fn=lambda x: tuple(zip(*x)))\n\nvalidation_dataset = SatoriusDataset(transforms=get_transform(train=False),root =root,train = False)\nvalidation_dataloader = DataLoader(validation_dataset, batch_size=1, shuffle=False, collate_fn=lambda x: tuple(zip(*x)))","metadata":{"scrolled":true,"execution":{"iopub.status.busy":"2023-03-30T17:30:08.239048Z","iopub.execute_input":"2023-03-30T17:30:08.239311Z","iopub.status.idle":"2023-03-30T17:30:09.257857Z","shell.execute_reply.started":"2023-03-30T17:30:08.239275Z","shell.execute_reply":"2023-03-30T17:30:09.257079Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"#### 2.Visualizing Dataset <a class=\"anchor\" id=\"3.2\"></a>","metadata":{}},{"cell_type":"code","source":"train_df = train_dataset.data_info\nvalidation_df = validation_dataset.data_info\n\npd.set_option('max_colwidth', 125)\ndisplay(train_df.head())\ndisplay(validation_df)","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:09.259913Z","iopub.execute_input":"2023-03-30T17:30:09.260342Z","iopub.status.idle":"2023-03-30T17:30:09.290930Z","shell.execute_reply.started":"2023-03-30T17:30:09.260303Z","shell.execute_reply":"2023-03-30T17:30:09.290218Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sns.set_palette(\"pastel\")\n\nn_annotations = [[],[]] #[[train], [validation]]\nfor i in train_df['annotation']:\n    n_annotations[0].append(len(i))\n    \nfor i in validation_df['annotation']:\n    n_annotations[1].append(len(i))\n\nx_axis,y_axis = 'Number Of Annotations' , 'Number Of Images'\n\nfig = plt.figure(figsize=(8,8))\np = sns.histplot(data = n_annotations[0])\np.set_xlabel(x_axis, fontsize = 15)\np.set_ylabel(y_axis, fontsize = 15)\n\n\nplt.show()\n\nfig = plt.figure(figsize=(8,8))\np = sns.histplot(data = n_annotations[1])\np.set_xlabel(x_axis, fontsize = 15)\np.set_ylabel(x_axis, fontsize = 15)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:09.292133Z","iopub.execute_input":"2023-03-30T17:30:09.293483Z","iopub.status.idle":"2023-03-30T17:30:09.982561Z","shell.execute_reply.started":"2023-03-30T17:30:09.293442Z","shell.execute_reply":"2023-03-30T17:30:09.981808Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import cv2\n\nimg, targets = next(iter(validation_dataloader))\nimage = np.array(img[0])\ntargets = targets[0]\nimage = cv2.cvtColor(image.transpose((1,2,0)), cv2.COLOR_BGR2RGB)\nall_masks = np.zeros((IMG_W ,IMG_H))\nfor mask in targets['masks']:\n    all_masks = np.logical_or(all_masks, mask)\n    \nplt.figure(figsize=(10, 10))\nplt.imshow(image)\nplt.axis(\"off\")\nplt.figure(figsize=(10, 10))\nplt.imshow(image)\nplt.imshow(all_masks, alpha=0.5)\nplt.axis(\"off\")\nplt.figure(figsize=(10, 10))\nplt.imshow(all_masks)\nplt.axis(\"off\")\n    \nplt.show()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:09.983993Z","iopub.execute_input":"2023-03-30T17:30:09.984442Z","iopub.status.idle":"2023-03-30T17:30:14.699510Z","shell.execute_reply.started":"2023-03-30T17:30:09.984400Z","shell.execute_reply":"2023-03-30T17:30:14.698794Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def clustered_img(x):\n    kmeans = cluster.KMeans(2)\n    dims = np.shape(x)\n    pixel_matrix = np.reshape(x, (dims[0] * dims[1], dims[2]))\n    clustered = kmeans.fit_predict(pixel_matrix)\n    clustered_img = np.reshape(clustered, (dims[0], dims[1]))\n    return clustered_img\n\nfor i in range(2):\n    fig = plt.figure(figsize=(10,10))\n    img, targets = next(iter(validation_dataloader))\n    img = img[0]\n    targets = targets[0]\n    plt.imshow(clustered_img(img.numpy().transpose((1,2,0))))\n    plt.grid(None)\n    plt.title(\"K-Means\")\n    plt.show()\n\n    fig = plt.figure(figsize=(10,10))\n    all_masks = np.zeros((IMG_W ,IMG_H))\n    for mask in targets['masks']:\n        all_masks = np.logical_or(all_masks, mask)\n    plt.imshow(img.numpy().transpose((1,2,0)))\n    plt.imshow(all_masks, alpha=0.3)\n    plt.grid(None)\n    plt.title(\"Target\")\n    plt.show()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:14.700790Z","iopub.execute_input":"2023-03-30T17:30:14.701181Z","iopub.status.idle":"2023-03-30T17:30:24.995847Z","shell.execute_reply.started":"2023-03-30T17:30:14.701143Z","shell.execute_reply":"2023-03-30T17:30:24.995218Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **4.Initializing pre-trained model** <a class=\"anchor\" id=\"4\"></a>","metadata":{}},{"cell_type":"code","source":"!mkdir -p /root/.cache/torch/hub/checkpoints/\n!cp ../input/cocopre/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth /root/.cache/torch/hub/checkpoints/maskrcnn_resnet50_fpn_coco-bf2d0c1e.pth","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:24.997185Z","iopub.execute_input":"2023-03-30T17:30:24.997775Z","iopub.status.idle":"2023-03-30T17:30:30.707815Z","shell.execute_reply.started":"2023-03-30T17:30:24.997735Z","shell.execute_reply":"2023-03-30T17:30:30.706700Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def get_model_instance_segmentation(num_classes):\n    # load an instance segmentation model pre-trained on COCO\n    model = torchvision.models.detection.maskrcnn_resnet50_fpn(pretrained=True)\n\n    # get number of input features for the classifier\n    in_features = model.roi_heads.box_predictor.cls_score.in_features\n    # replace the pre-trained head with a new one\n#     model.roi_heads.box_predictor = FastRCNNPredictor(in_features, num_classes)  #If to train newly\n\n    # now get the number of input features for the mask classifier\n    in_features_mask = model.roi_heads.mask_predictor.conv5_mask.in_channels\n    hidden_layer = 256\n    # and replace the mask predictor with a new one\n    model.roi_heads.mask_predictor = MaskRCNNPredictor(in_features_mask,\n                                                       hidden_layer,\n                                                       num_classes)\n\n    return model","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:30.709671Z","iopub.execute_input":"2023-03-30T17:30:30.709958Z","iopub.status.idle":"2023-03-30T17:30:30.717411Z","shell.execute_reply.started":"2023-03-30T17:30:30.709922Z","shell.execute_reply":"2023-03-30T17:30:30.716424Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')\n    # our dataset has two classes only - background and cell\nnum_classes = 2\n\nmodel = get_model_instance_segmentation(num_classes)\n\n    # move model to the right device\nmodel.to(device)\n\n    # construct an optimizer\nparams = [p for p in model.parameters() if p.requires_grad]\noptimizer = torch.optim.SGD(params, lr=LEARNING_RATE, momentum=MOMENTUM, weight_decay=WEIGHT_DECAY)\nepoch = 25\nloss_history = [[],[]]\ntrain_n_minibatches = train_dataloader.__len__()\nvalidation_n_minibatches = validation_dataloader.__len__()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:30.721277Z","iopub.execute_input":"2023-03-30T17:30:30.721817Z","iopub.status.idle":"2023-03-30T17:30:34.565070Z","shell.execute_reply.started":"2023-03-30T17:30:30.721780Z","shell.execute_reply":"2023-03-30T17:30:34.564221Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# WIDTH = 704\n# HEIGHT = 520\n# DEVICE = 'cuda'\n\n# def analyze_train_sample(model, ds_train, sample_index):\n    \n#     img, targets = ds_train[sample_index]\n#     plt.imshow(img.numpy().transpose((1,2,0)))\n#     plt.title(\"Image\")\n#     plt.show()\n    \n#     masks = np.zeros((HEIGHT, WIDTH))\n#     for mask in targets['masks']:\n#         masks = np.logical_or(masks, mask)\n#     plt.imshow(img.numpy().transpose((1,2,0)))\n#     plt.imshow(masks, alpha=0.3)\n#     plt.title(\"Ground truth\")\n#     plt.show()\n    \n#     model.eval()\n#     with torch.no_grad():\n#         preds = model([img.to(DEVICE)])[0]\n\n#     plt.imshow(img.cpu().numpy().transpose((1,2,0)))\n#     all_preds_masks = np.zeros((HEIGHT, WIDTH))\n#     for mask in preds['masks'].cpu().detach().numpy():\n#         all_preds_masks = np.logical_or(all_preds_masks, mask[0] > MASK_THRESHOLD)\n#     plt.imshow(all_preds_masks, alpha=0.4)\n#     plt.title(\"Predictions\")\n#     plt.show()\n    \n# analyze_train_sample(model, train_dataset, 20)","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:34.566501Z","iopub.execute_input":"2023-03-30T17:30:34.566754Z","iopub.status.idle":"2023-03-30T17:30:41.766040Z","shell.execute_reply.started":"2023-03-30T17:30:34.566719Z","shell.execute_reply":"2023-03-30T17:30:41.764652Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **5.Training** <a class=\"anchor\" id=\"5\"></a>","metadata":{}},{"cell_type":"code","source":"model.train()\nlog_idx = 100\nimport time\ntotal_time = 0\nstart = time.time()\n\n\nfor e in range(epoch):\n    for batch_idx , (x ,y) in enumerate(train_dataloader):\n        optimizer.zero_grad()\n        x = list(image.to(device) for image in x)\n        y = [{k: v.to(device) for k, v in t.items()} for t in y]\n        loss_dict = model(x, y)\n        \n        loss = sum(loss for loss in loss_dict.values())\n        loss.backward()\n        optimizer.step()\n        loss_history[0].append(float(loss.detach()))\n        total_time += time.time() - start\n        start = time.time()\n        \n        if batch_idx % log_idx == 0:\n            # Printing Log\n            print(f'LOSS for EPOCH {e+1} BATCH {batch_idx+1}/{train_n_minibatches} TRAIN LOSS : {loss_history[0][-1]}',end = ' ')\n            with torch.no_grad():\n                # Calculating loss and accuracy for validation\n                for _batch_idx_ , (x ,y) in enumerate(validation_dataloader):\n                    x = list(image.to(device) for image in x)\n                    y = [{k: v.to(device) for k, v in t.items()} for t in y]\n                    loss_dict = model(x, y)\n                    validation_loss = sum(loss for loss in loss_dict.values())\n                    loss_history[1].append(float(validation_loss.detach()))\n                                      \n                print(f'VALIDATION LOSS : {sum(loss_history[1][-1:-validation_n_minibatches-1:-1])/validation_n_minibatches}')\n\n    torch.save(model.state_dict(),'masked_rcnn_ss')\n    #Log for e+1th epoch\n    print(f'---------------------------------------EPOCH {e+1}-------------------------------------------')\n    print(f'Loss for EPOCH {e+1}  TRAIN LOSS : {sum(loss_history[0][-1:-train_n_minibatches-1:-1])/train_n_minibatches}')\n    n_validation_losses = int(train_n_minibatches/log_idx)*validation_n_minibatches\n    print(f'VALIDATION LOSS for EPOCH {e+1} : {sum(loss_history[1][-1:-1*n_validation_losses-1:-1])/n_validation_losses}',end = '\\n')\n    print('---------------------------------------------------------------------------------------------')","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.767447Z","iopub.status.idle":"2023-03-30T17:30:41.767882Z","shell.execute_reply.started":"2023-03-30T17:30:41.767652Z","shell.execute_reply":"2023-03-30T17:30:41.767677Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(f'Total time taken for training : {total_time} seconds')","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.769165Z","iopub.status.idle":"2023-03-30T17:30:41.770161Z","shell.execute_reply.started":"2023-03-30T17:30:41.769870Z","shell.execute_reply":"2023-03-30T17:30:41.769898Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **6.Plotting Graphs** <a class=\"anchor\" id=\"6\"></a>","metadata":{}},{"cell_type":"markdown","source":"# 1.Plotting Loss vs Epoch<a class=\"anchor\" id=\"6.1\"></a>","metadata":{}},{"cell_type":"code","source":"# Plotting Loss per epoch\nloss_per_epoch = [[],[]]\nfor i in range(epoch):\n    temp = 0\n    for j in loss_history[0][i*train_n_minibatches:(i+1)*train_n_minibatches]:\n        temp = temp + j\n    loss_per_epoch[0].append(temp/train_n_minibatches)\n    temp = 0\n    for j in loss_history[1][i*n_validation_losses:(i+1)*n_validation_losses]:\n        temp = temp + j\n    loss_per_epoch[1].append(temp/n_validation_losses)    \n\nsns.lineplot(x=range(len(loss_per_epoch[0])),y=loss_per_epoch[0])\nsns.lineplot(x=range(len(loss_per_epoch[1])),y=loss_per_epoch[1])\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.771329Z","iopub.status.idle":"2023-03-30T17:30:41.772279Z","shell.execute_reply.started":"2023-03-30T17:30:41.772038Z","shell.execute_reply":"2023-03-30T17:30:41.772065Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **7.Loading and Testing**<a class=\"anchor\" id=\"7\"></a>","metadata":{}},{"cell_type":"code","source":"model = get_model_instance_segmentation(num_classes)\n    # move model to the right device\nmodel.to(device)\nmodel.load_state_dict(torch.load('masked_rcnn_ss', map_location='cuda'))","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.773376Z","iopub.status.idle":"2023-03-30T17:30:41.774390Z","shell.execute_reply.started":"2023-03-30T17:30:41.774106Z","shell.execute_reply":"2023-03-30T17:30:41.774132Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_model(model, validation_dataloader,n=3)","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.775517Z","iopub.status.idle":"2023-03-30T17:30:41.776367Z","shell.execute_reply.started":"2023-03-30T17:30:41.776102Z","shell.execute_reply":"2023-03-30T17:30:41.776127Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### **8.Predictions**<a class=\"anchor\" id=\"8\"></a>","metadata":{}},{"cell_type":"code","source":"class TestDataset(Dataset):\n    def __init__(self, root,transforms = None):\n        self.root = root\n        self.transforms = transforms\n        self.img_name = []\n        for i in os.listdir(root+'test/'):\n            self.img_name.append(i[:-4])\n        self.w , self.h = 520 , 704\n\n            \n    def __getitem__(self, idx):\n        image = Image.open(self.root+'test/'+self.img_name[idx]+'.png').convert(\"RGB\")\n        if self.transforms is not None:\n            image, _ = self.transforms(image=image, target=None)\n        \n        return image , self.img_name[idx]\n\n    def __len__(self):\n        return len(self.img_name)\n    \n","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.777664Z","iopub.status.idle":"2023-03-30T17:30:41.778492Z","shell.execute_reply.started":"2023-03-30T17:30:41.778220Z","shell.execute_reply":"2023-03-30T17:30:41.778245Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"td = TestDataset(root, transforms=get_transform(train=False))","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.779695Z","iopub.status.idle":"2023-03-30T17:30:41.780541Z","shell.execute_reply.started":"2023-03-30T17:30:41.780273Z","shell.execute_reply":"2023-03-30T17:30:41.780298Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"DROP_OVERLAPPING = False\n\nsubmission = []\ncounter = 0\n\nmin_precision = 0.5\n\nmodel.eval()\n\nfor img,img_name in td:\n    \n    with torch.no_grad():\n        result = model([img.to(device)])[0]\n        \n    if len(result[\"masks\"]) != 0:\n        previous_masks = []\n        for j, m in enumerate(result[\"masks\"]):\n            original_mask = result[\"masks\"][j][0].cpu().numpy()\n            if DROP_OVERLAPPING and does_overlap(original_mask, previous_masks):\n                continue\n            else:\n                original_mask = remove_overlapping_pixels(original_mask, previous_masks)\n                previous_masks.append(original_mask)\n                rle = rle_encoding(original_mask > min_precision)\n                submission.append([img_name, rle])\n    else:\n        submission.append([img_name, \"\"])\n\ndf_sub = pd.DataFrame(submission, columns=['id', 'predicted'])\ndf_sub.to_csv(\"submission.csv\", index=False)\ndf_sub.head()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.781774Z","iopub.status.idle":"2023-03-30T17:30:41.782715Z","shell.execute_reply.started":"2023-03-30T17:30:41.782469Z","shell.execute_reply":"2023-03-30T17:30:41.782502Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_sub.groupby('id')['predicted'].agg(lambda x: list(x)).reset_index()","metadata":{"execution":{"iopub.status.busy":"2023-03-30T17:30:41.783859Z","iopub.status.idle":"2023-03-30T17:30:41.784866Z","shell.execute_reply.started":"2023-03-30T17:30:41.784608Z","shell.execute_reply":"2023-03-30T17:30:41.784633Z"},"trusted":true},"execution_count":null,"outputs":[]}]}