{"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":"This is notebook is based on [CassavaLeaf ViT baseline Notebook](https://www.kaggle.com/szuzhangzhi/vision-transformer-vit-cuda-as-usual) and Adjusted for Plant pathology 2021 dataset.","metadata":{}},{"cell_type":"markdown","source":"Обычно картинку используют как 3D array (высота, ширина, количество каналов) и применяется к ним сверточные слои. Но тут есть ряд недостатков:\n\n- не все пиксели одинаково полезны;\n- свертки не достаточно хорошо работают с пикселями, находящимися далеко друг от друга;\n- свертки недостаточно эффективны в очень глубоких нейронных сетях.\n\nВ результате авторы предлагают конвертировать изображения в визуальные токены и подавать их в трансформер.\n\n- Вначале используется обычный backbone для получения feature maps\n- Далее feature map конвертируется в визуальные токены\n- Токены подаются в трансформеры\n- Выход трансформера может использоваться для задач классификации\n- А если объединить выход трансформера с feature map, то можно получить предсказания для задач сегментации","metadata":{}},{"cell_type":"markdown","source":"Self-Attention между пикселями. К примеру, если картинка размера 640x640, модельке надо посчитать self-attention для 409к комбинаций. Но вероятней всего, самый верхний правый пиксель вряд ли будет иметь значимое влияние на нижний левый пиксель. ViT справился с этой проблемой за счет сегментации картинки на маленькие патчи (к примеру, 16x16).\n\n![image.png](attachment:cded490a-0e7b-4b0e-99c5-cd91584bb995.png)","metadata":{},"attachments":{"cded490a-0e7b-4b0e-99c5-cd91584bb995.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"1. Изображение нужно поделить на пачти фиксированного размера\n\nЕсть 2D картинка размера  H * W, которая может разделиться на N патчей, где $N=\\frac{H * W}{P^2}$. Если картинка 48x48, размер патча 16x16, то получится $N=\\frac{48 * 48}{16^2} = 9$.","metadata":{}},{"cell_type":"code","source":"from scipy import misc\nimport matplotlib.pyplot as plt\n\n\nimg = misc.face()\nplt.imshow(img);","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import cv2\nimg = cv2.resize(img, (224, 224))\nplt.imshow(img);","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(1, 14)\nfor i in range(14):\n#     for j in range(5):\n    patch = img[0:16, 16 * i:16 * (i + 1)]\n    ax[i].imshow(patch)\n    ax[i].axis('off')\nplt.show()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n2. Вытянуть 2D патчи в 1D представление\n\nКаждый патч вытягивается в 1D патч путем конкатенации всех пикселей и затем применяется линейная проекция до желаемой размерности - это будет пачт эмбеддинга. (Эта матрица перевода для всех патчей одинаковая)","metadata":{}},{"cell_type":"code","source":"16 * 16","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"patch.flatten().shape","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"\n3. Позиционный эмбеддинг добавляется к патчу эмбэддинг, чтобы сохранить позиционную информацию\n\nЭто обучаемая таблица с позиционными векторами.\n\n\nТрансформеры не зависят от структуры входных элементов, поэтому добавление обучаемых позиционных эмбэддингов в каждый патч позволит модели узнать о структуре изображения.\n\n\n\nЭта последовательность векторов патчей будет использоваться в качестве входной последовательности для энкодера трансформера.","metadata":{}},{"cell_type":"markdown","source":"### Transformer Encoder\n\n![image.png](attachment:6a08e9ff-e266-493e-880d-5ab34014dc73.png)\n\nЭнкодер состоит из\n- Multi-Head Self Attention Layer(MSP) для линейного объединения нескольких выходов внимания в соответствии с ожидаемыми размерами. Несколько Multi-Head помогают изучать локальные и глобальные зависимости в изображении.\n- Многослойные персептроны (MLP) содержат два слоя с GELU (Gaussian Error Linear Unit)\n- Layer Norm (LN) применяется перед каждым блоком, так как она не вводит никаких новых зависимостей между обучающими изображениями. Помогает улучшить время обучения и производительность обобщения\n- Residual connections применяются после каждого блока\n\nДля классификации изображений классификация реализована с использованием MLP с одним скрытым слоем.\nВерхние слои ViT изучают глобальные объекты, тогда как нижние слои изучают как глобальные, так и локальные объекты. Это позволяет ViT изучать более общие паттерны.\n","metadata":{},"attachments":{"6a08e9ff-e266-493e-880d-5ab34014dc73.png":{"image/png":"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"}}},{"cell_type":"markdown","source":"### Self-Attention\nБудем работать со следующим предложением:\n\n”The animal didn't cross the street because it was too tired”\n\nЧто означает it в этом предложении? It относится к улице или к животному? Когда модель обрабатывает слово it, self-attention позволяет ей ассоциировать it с animal, но it еще и связано с tired.\n\nSelf-attention позволяет нейросетке смотреть на другие слова во входной последовательности в поисках подсказок, которые могут помочь привести к лучшему кодированию этого слова.\n\n\n<img src='http://jalammar.github.io/images/t/transformer_self-attention_visualization.png' width=400>","metadata":{}},{"cell_type":"markdown","source":"\nПервый шаг в вычислении self-attention - создание трех векторов из каждого входного эмбеддинга. Для каждого слова создается вектор Query, вектор Key и вектор Value. Эти векторы создаются путем умножения эмбеддинга на три матрицы, которые обучаются.\n- Query - слово, с которого смотрим на всё остальное\n- Key - слово, на которое смотрим\n- Value - здесь содержится смысл слова\n\nПри перемножении Query на Key мы понимаем насколько слово с Key релевантно слову с Query.\n\n<img src='http://jalammar.github.io/images/t/transformer_self_attention_vectors.png' width=500>\n\n\nДеление на $\\sqrt{d}$ (квадратный корень из размерности Key векторов. для того, чтобы были более стабильные градиенты.\n\n<img src='http://jalammar.github.io/images/t/self-attention-matrix-calculation-2.png' width=400>","metadata":{}},{"cell_type":"markdown","source":"### Multi-Head Self-Attention\n\nЭто улучшает работу слоя внимания двумя способами:\n\n1. Это расширяет возможности модели фокусироваться на различных словах.\n\n2. Это дает слою внимания несколько “подпространств представления”. С Multi-Head self-attention есть несколько наборов весовых матриц query/key/value. Каждый из этих наборов инициализируется случайным образом. Затем, после обучения, каждый набор используется для проецирования входных вложений в другое подпространство представления.\n\n<img src='http://jalammar.github.io/images/t/transformer_attention_heads_qkv.png' width=600>","metadata":{}},{"cell_type":"markdown","source":"Ссылки:\n1. [Vision Transformers: A New Computer Vision Paradigm](https://medium.com/swlh/visual-transformers-a-new-computer-vision-paradigm-aa78c2a2ccf2)\n2. [Visual Transformers: Token-based Image Representation and Processing for Computer Vision](https://arxiv.org/pdf/2006.03677.pdf)\n3. [Vision Transformer (ViT) - An image is worth 16x16 words | Paper Explained](https://www.youtube.com/watch?v=j6kuz_NqkG0)\n4. [Прикладное машинное обучение 4. Self-Attention. Transformer overview](https://youtu.be/UETKUIlYE6g)","metadata":{}},{"cell_type":"code","source":"import sys\n\npackage_path = '../input/vision-transformer-pytorch/VisionTransformer-Pytorch'\nsys.path.append(package_path)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"import os\nimport pandas as pd\n\nimport time\nimport datetime\nimport copy\nimport matplotlib.pyplot as plt\nimport json\nimport seaborn as sns\nimport cv2\nimport albumentations as albu\nimport numpy as np\n\nimport torch\nimport torch.nn as nn\nimport torchvision.models as models\nimport torch.optim as optim\nfrom torch.utils.data import Dataset, DataLoader\nfrom torch.optim.lr_scheduler import ReduceLROnPlateau\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.model_selection import StratifiedKFold, GroupKFold, KFold, train_test_split\n\n\n# ALBUMENTATIONS\nimport albumentations as albu\n\nfrom albumentations import (\n    HorizontalFlip, VerticalFlip, IAAPerspective, ShiftScaleRotate, CLAHE, RandomRotate90,\n    Transpose, ShiftScaleRotate, Blur, OpticalDistortion, GridDistortion, HueSaturationValue,\n    IAAAdditiveGaussianNoise, GaussNoise, MotionBlur, MedianBlur, IAAPiecewiseAffine, RandomResizedCrop,\n    IAASharpen, IAAEmboss, RandomBrightnessContrast, Flip, OneOf, Compose, Normalize, Cutout, CoarseDropout, ShiftScaleRotate, CenterCrop, Resize\n)\n    \nfrom albumentations.pytorch import ToTensorV2","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# BASE_DIR=\"../input/plant-pathology-2021-fgvc8/\"\nBASE_DIR = '../input/plant-pathology-2021-224x224/'\nTRAIN_IMAGES_DIR = os.path.join(BASE_DIR, 'train_imgs')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_df = pd.read_csv('../input/plant-pathology-2021-fgvc8/train.csv')\ntrain_df.head()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print(\"Count of training images {0}\".format(len(os.listdir(TRAIN_IMAGES_DIR))))","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"class_name = train_df['labels'].value_counts().index\nclass_count = train_df['labels'].value_counts().values","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_df.labels.value_counts()","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plt.figure(figsize=(8,5))\nsns.countplot(data=train_df, y='labels');","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from sklearn.preprocessing import LabelEncoder\nle = LabelEncoder()\n\nle.fit(train_df.labels)\ntrain_df['labels'] = le.transform(train_df.labels)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def visualize_images(image_ids, labels):\n    plt.figure(figsize=(16, 12))\n    \n    for idx, (image_id, label) in enumerate(zip(image_ids, labels)):\n        plt.subplot(3, 3, idx+1)\n        \n        image = cv2.imread(os.path.join(TRAIN_IMAGES_DIR, image_id))\n        image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)\n        \n        plt.imshow(image)\n        plt.title(f\"Class: {label}\", fontsize=12)\n        \n        plt.axis(\"off\")\n        \n    plt.show()\n    \n\ndef plot_augmentation(image_id, transform):\n    plt.figure(figsize=(16, 4))\n    \n    img = cv2.imread(os.path.join(TRAIN_IAMGES_DIR, image_id))\n    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n    \n    plt.subplot(1, 3, 1)\n    plt.imshow(img)\n    plt.axis('off')\n    \n    plt.subplot(1, 3, 2)\n    x = transform(image=img)['image']\n    plt.imshow(x)\n    plt.axis('off')\n    \n    plt.subplot(1, 3, 3)\n    x = transform(image=img)['image']\n    plt.imshow(x)\n    \ndef visualize(images, transform):\n    '''\n    Plot images and their transformations\n    '''\n    fig = plt.figure(figsize=(32, 16))\n    \n    for i, im in enumerate(images):\n        ax = fig.add_subplot(2, 5, i+1, xticks=[], yticks=[])\n        plt.imshow(im)\n        \n    for i, im in enumerate(images):\n        ax = fig.add_subplot(2, 5, i+6, xticks=[], yticks=[])\n        plt.imshow(transform(image=im)['image'])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# CUSTOM DATASET CLASS\nclass PlantDataset(Dataset):\n    def __init__(\n        self, df:pd.DataFrame, imfolder:str, train:bool=True, transforms=None\n    ):\n        self.df = df\n        self.imfolder = imfolder\n        self.train = train\n        self.transforms = transforms\n        \n    def __len__(self):\n        return len(self.df)\n    \n    def __getitem__(self, index):\n        im_path = os.path.join(self.imfolder, self.df.iloc[index]['image'])\n        im = cv2.imread(im_path, cv2.IMREAD_COLOR)\n        im = cv2.cvtColor(im, cv2.COLOR_BGR2RGB)\n        \n        if (self.transforms):\n            '''\n            When AlbumentationCompose, a dictionary with key 'image' is created\n            '''\n            im = self.transforms(image=im)['image']\n            \n        if (self.train):\n            label = self.df.iloc[index]['labels']\n            return im, label\n        else:\n            return im","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# AUGMENTATIONS\n\ntrain_augs = albu.Compose([\n    albu.RandomResizedCrop(height=384, width=384, p=1.0),\n    albu.HorizontalFlip(p=0.5),\n    albu.VerticalFlip(p=0.5),\n    albu.RandomBrightnessContrast(p=0.5),\n    albu.ShiftScaleRotate(p=0.5),\n    albu.Normalize(    \n        mean=[0.3, 0.3, 0.3],\n        std=[0.1, 0.1, 0.1],),\n    CoarseDropout(p=0.5),\n    Cutout(p=0.5),\n    ToTensorV2(),\n])\n\nvalid_augs = albu.Compose([\n    albu.Resize(height=384, width=384, p=1.0),\n    albu.Normalize(\n        mean=[0.3, 0.3, 0.3],\n        std=[0.1, 0.1, 0.1],),\n    ToTensorV2(),\n])\n","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# DATA SPLIT\ntrain, valid = train_test_split(\n    train_df,\n    test_size=0.1,\n    random_state=42,\n    stratify=train_df.labels.values\n)\n\n# reset index on both dataframes\ntrain = train.reset_index(drop=True)\nvalid = valid.reset_index(drop=True)\n\n# targets in train,valid datasets\ntrain_targets = train.labels.values\nvalid_targets = valid.labels.values","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# DEFINE PYTORCH CUSTOM DATASET\ntrain_dataset = PlantDataset(\n    df=train,\n    imfolder=TRAIN_IMAGES_DIR,\n    train=True,\n    transforms=train_augs\n)\n\nvalid_dataset = PlantDataset(\n    df=valid,\n    imfolder=TRAIN_IMAGES_DIR,\n    train=True,\n    transforms=valid_augs\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def plot_image(img_dict):\n    image_tensor = img_dict[0]\n    target = img_dict[1]\n    print(target)\n    image = image_tensor.permute(1, 2, 0)\n    plt.imshow(image)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"plot_image(train_dataset[7])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# MAKE PYTORCH DATALOADER\ntrain_loader = DataLoader(\n    train_dataset,\n    batch_size=8,\n    num_workers=4,\n    shuffle = True\n)\n\nvalid_loader = DataLoader(\n    valid_dataset,\n    batch_size=8,\n    num_workers=4,\n    shuffle = False\n)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# TRAIN\ndef train_model(datasets, dataloaders, model, criterion, optimizer, scheduler, num_epochs, device):\n    since = time.time()\n    \n    best_model_wts = copy.deepcopy(model.state_dict())\n    best_acc = 0.0\n    \n    for epoch in range(num_epochs):\n        print('Epoch {}/{}'.format(epoch, num_epochs-1))\n        print('-' * 10)\n        \n        for phase in ['train', 'valid']:\n            if phase == 'train':\n                model.train()\n            else:\n                model.eval()\n                \n            running_loss = 0.0\n            running_corrects = 0.0\n            running_total = 0.0\n            \n            for step, (inputs, labels) in enumerate(dataloaders[phase]):\n                inputs = inputs.to(device)\n                labels = labels.to(device)\n               \n                # Zero out the grads\n                optimizer.zero_grad()\n                \n                # Forward\n                # Track history in train mode\n                with torch.set_grad_enabled(phase == 'train'):\n                    model = model.to(device)\n                    outputs = model(inputs)\n                    _, preds = torch.max(outputs, 1) \n                    loss = criterion(outputs, labels)\n                    \n                    if phase == 'train':\n                        loss.backward()\n                        optimizer.step()\n                \n                # Statistics\n                running_loss += loss.item() * inputs.size(0)\n                running_corrects += torch.sum(preds == labels.data)\n                running_total += len(labels.data)\n                \n                if (step + 1) % 100 == 0:\n                    print(f'[{step + 1}/{len(dataloaders[phase])}].')\n                    print(f'Loss {running_loss / running_total}. Accuracy {running_corrects / running_total}')\n            \n            if phase == 'train':\n                scheduler.step()\n                \n            epoch_loss = running_loss / len(datasets[phase])\n            epoch_acc = running_corrects.double() / len(datasets[phase])\n            \n            print('{} Loss: {:.4f} Acc: {:.4f}'.format(\n                phase, epoch_loss, epoch_acc))\n            \n            if phase == 'valid' and epoch_acc > best_acc:\n                best_acc = epoch_acc\n                best_model_wts = copy.deepcopy(model.state_dict())\n        \n        print()\n    \n    time_elapsed = time.time() - since\n    print('Training complete in {:.0f}m {:.0f}s'.format(time_elapsed // 60, time_elapsed % 60))\n    print('Best val Acc: {:.4f}'.format(best_acc))\n    \n    model.load_state_dict(best_model_wts)\n    \n    return model","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"### Training","metadata":{}},{"cell_type":"code","source":"# from vision_transformer_pytorch import VisionTransformer\n# device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n# print(device)\n\n# datasets = {'train': train_dataset,\n#             'valid': valid_dataset}\n\n# dataloaders = {'train': train_loader,\n#                'valid': valid_loader}\n\n# # # LOAD PRETRAINED ViT MODEL\n# # model = VisionTransformer.from_pretrained('ViT-B_16', num_classes=12)        \n\n# model_path = '../input/plantpathologyvitb169epochs/vit_b-16_9epoch_pretrained.pt'\n# model = torch.load(model_path)\n\n# # OPTIMIZER\n# optimizer = torch.optim.AdamW(model.parameters(), lr=1e-4, weight_decay=0.001)\n# # optimizer = torch.optim.Adam(model.parameters(), lr=1e-4, weight_decay=0.001)\n# # optimizer = AdamP(model.parameters(), lr=1e-4, weight_decay=0.001)\n\n# # LEARNING RATE SCHEDULER\n# scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=2, gamma=0.1)\n\n# criterion = nn.CrossEntropyLoss()\n# num_epochs = 3","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# torch.save(model, 'full_plant_pathology_vit.pt')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# len(dataloaders['train'])","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# # MODEL TRAIN\n# trained_model = train_model(datasets, dataloaders,\n#                             model, criterion,\n#                             optimizer, scheduler,\n#                             num_epochs, device)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Save the mode after training\n# torch.save(model, 'vit_b-16_12epoch_pretrained.pt')","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"## Testing","metadata":{}},{"cell_type":"code","source":"from vision_transformer_pytorch import VisionTransformer\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\nprint(device)\n\n# LOAD PRETRAINED ViT MODEL\nmodel_path = '../input/plantpathologyvitb169epochs/vit_b-16_12epoch_pretrained.pt'\nmodel = torch.load(model_path)\n# model.load_state_dict(torch.load(PATH))\nmodel.eval()","metadata":{"_kg_hide-output":true,"scrolled":true,"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"from pathlib import Path\n\ndf_preds = pd.DataFrame()\nfor path in Path('../input/plant-pathology-2021-fgvc8/test_images').iterdir():\n    img = cv2.imread(str(path))\n    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)\n    img = valid_augs(image=img)['image'].cuda()\n    pred = model(img[None])\n    \n    df_preds = df_preds.append(\n        {'image': path.parts[-1], 'labels': le.inverse_transform(torch.argmax(pred.cpu(), dim=1))[0]},\n        ignore_index=True)\n    \ndf_preds","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_preds.to_csv('submission.csv', index=False)","metadata":{"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"Полезные ссылки:\n1. [Swin-Transformer](https://github.com/microsoft/Swin-Transformer)\n2. [VisionTransformer-Pytorch](https://github.com/tczhangzhi/VisionTransformer-PyTorch)\n3. [vit-pytorch](https://github.com/lucidrains/vit-pytorch)","metadata":{}}]}