102. Kidney_segmentation_with_Unet.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[],"collapsed_sections":[],"mount_file_id":"196krwAugUz5GPwFhUD-tbeoADseyFOvZ","authorship_tag":"ABX9TyPeBnb6HxOYos5oizPZAsGF"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"}},"cells":[{"cell_type":"code","execution_count":null,"metadata":{"id":"NBD5-OzuuQne","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1664112528014,"user_tz":-360,"elapsed":4292,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"7fc24948-0dc9-4181-9515-155adbe50b6a"},"outputs":[{"output_type":"stream","name":"stdout","text":["Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n","Requirement already satisfied: tensorflow in /usr/local/lib/python3.7/dist-packages (2.8.2+zzzcolab20220719082949)\n","Requirement already satisfied: libclang>=9.0.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (14.0.6)\n","Requirement already satisfied: google-pasta>=0.1.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (0.2.0)\n","Requirement already satisfied: flatbuffers>=1.12 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (2.0.7)\n","Requirement already satisfied: wrapt>=1.11.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.14.1)\n","Requirement already satisfied: protobuf<3.20,>=3.9.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (3.17.3)\n","Requirement already satisfied: h5py>=2.9.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (3.1.0)\n","Requirement already satisfied: astunparse>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.6.3)\n","Requirement already satisfied: numpy>=1.20 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.21.6)\n","Requirement already satisfied: tensorboard<2.9,>=2.8 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (2.8.0)\n","Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (0.26.0)\n","Requirement already satisfied: gast>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (0.5.3)\n","Requirement already satisfied: absl-py>=0.4.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.2.0)\n","Requirement already satisfied: tensorflow-estimator<2.9,>=2.8 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (2.8.0)\n","Requirement already satisfied: setuptools in /usr/local/lib/python3.7/dist-packages (from tensorflow) (57.4.0)\n","Requirement already satisfied: keras<2.9,>=2.8.0rc0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (2.8.0)\n","Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.1.0)\n","Requirement already satisfied: grpcio<2.0,>=1.24.3 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.48.1)\n","Requirement already satisfied: typing-extensions>=3.6.6 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (4.1.1)\n","Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (3.3.0)\n","Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.15.0)\n","Requirement already satisfied: keras-preprocessing>=1.1.1 in /usr/local/lib/python3.7/dist-packages (from tensorflow) (1.1.2)\n","Requirement already satisfied: wheel<1.0,>=0.23.0 in /usr/local/lib/python3.7/dist-packages (from astunparse>=1.6.0->tensorflow) (0.37.1)\n","Requirement already satisfied: cached-property in /usr/local/lib/python3.7/dist-packages (from h5py>=2.9.0->tensorflow) (1.5.2)\n","Requirement already satisfied: google-auth<3,>=1.6.3 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (1.35.0)\n","Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (0.4.6)\n","Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (2.23.0)\n","Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (1.8.1)\n","Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (3.4.1)\n","Requirement already satisfied: werkzeug>=0.11.15 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (1.0.1)\n","Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /usr/local/lib/python3.7/dist-packages (from tensorboard<2.9,>=2.8->tensorflow) (0.6.1)\n","Requirement already satisfied: cachetools<5.0,>=2.0.0 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (4.2.4)\n","Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (0.2.8)\n","Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.7/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (4.9)\n","Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.7/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow) (1.3.1)\n","Requirement already satisfied: importlib-metadata>=4.4 in /usr/local/lib/python3.7/dist-packages (from markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow) (4.12.0)\n","Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.7/dist-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard<2.9,>=2.8->tensorflow) (3.8.1)\n","Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /usr/local/lib/python3.7/dist-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.9,>=2.8->tensorflow) (0.4.8)\n","Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (1.24.3)\n","Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (2.10)\n","Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (2022.6.15)\n","Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests<3,>=2.21.0->tensorboard<2.9,>=2.8->tensorflow) (3.0.4)\n","Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.7/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.9,>=2.8->tensorflow) (3.2.0)\n"]}],"source":["!pip install tensorflow"]},{"cell_type":"markdown","source":["**Acknowledgements:**\n","\n","Codes are taken from [Idiot Developer](https://www.youtube.com/watch?v=LkmgdNpcgTM&t=841s) and modified according to the need for this project."],"metadata":{"id":"-k6SMlg6l4QM"}},{"cell_type":"code","source":["from tensorflow.keras.layers import Conv2D, BatchNormalization, Activation, MaxPool2D, Conv2DTranspose, Concatenate, Input\n","from tensorflow.keras.models import Model\n","\n","def conv_block(input, num_filters):\n","    x = Conv2D(num_filters, 3, padding=\"same\")(input)\n","    x = BatchNormalization()(x)\n","    x = Activation(\"relu\")(x)\n","\n","    x = Conv2D(num_filters, 3, padding=\"same\")(x)\n","    x = BatchNormalization()(x)\n","    x = Activation(\"relu\")(x)\n","\n","    return x\n","\n","def encoder_block(input, num_filters):\n","    x = conv_block(input, num_filters)\n","    p = MaxPool2D((2, 2))(x)\n","    return x, p\n","\n","def decoder_block(input, skip_features, num_filters):\n","    x = Conv2DTranspose(num_filters, (2, 2), strides=2, padding=\"same\")(input)\n","    x = Concatenate()([x, skip_features])\n","    x = conv_block(x, num_filters)\n","    return x\n","\n","def build_unet(input_shape):\n","    inputs = Input(input_shape)\n","\n","    s1, p1 = encoder_block(inputs, 64)\n","    s2, p2 = encoder_block(p1, 128)\n","    s3, p3 = encoder_block(p2, 256)\n","    s4, p4 = encoder_block(p3, 512)\n","\n","    b1 = conv_block(p4, 1024)\n","\n","    d1 = decoder_block(b1, s4, 512)\n","    d2 = decoder_block(d1, s3, 256)\n","    d3 = decoder_block(d2, s2, 128)\n","    d4 = decoder_block(d3, s1, 64)\n","\n","    outputs = Conv2D(1, 1, padding=\"same\", activation=\"sigmoid\")(d4)\n","\n","    model = Model(inputs, outputs, name=\"U-Net\")\n","    return model\n","\n","if __name__ == \"__main__\":\n","    input_shape = (512, 512, 3)\n","    \n","    model = build_unet(input_shape)\n","    #model.summary()"],"metadata":{"id":"TmmDurIauVrc"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["import numpy as np\n","import tensorflow as tf\n","from tensorflow.keras import backend as K\n","\n","def iou(y_true, y_pred):\n","    def f(y_true, y_pred):\n","        intersection = (y_true * y_pred).sum()\n","        union = y_true.sum() + y_pred.sum() - intersection\n","        x = (intersection + 1e-15) / (union + 1e-15)\n","        x = x.astype(np.float32)\n","        return x\n","    return tf.numpy_function(f, [y_true, y_pred], tf.float32)\n","\n","smooth = 1e-15\n","def dice_coef(y_true, y_pred):\n","    y_true = tf.keras.layers.Flatten()(y_true)\n","    y_pred = tf.keras.layers.Flatten()(y_pred)\n","    intersection = tf.reduce_sum(y_true * y_pred)\n","    return (2. * intersection + smooth) / (tf.reduce_sum(y_true) + tf.reduce_sum(y_pred) + smooth)\n","\n","def dice_loss(y_true, y_pred):\n","    return 1.0 - dice_coef(y_true, y_pred)"],"metadata":{"id":"2VuQPeixuZPw"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["import os\n","os.environ[\"TF_CPP_MIN_LOG_LEVEL\"] = \"2\"\n","import numpy as np\n","import cv2\n","from glob import glob\n","from sklearn.model_selection import train_test_split\n","import tensorflow as tf\n","from tensorflow.keras.callbacks import ModelCheckpoint, CSVLogger, ReduceLROnPlateau\n","from tensorflow.keras.optimizers import Adam\n","from tensorflow.keras.metrics import Recall, Precision\n"," \n","\n","\"\"\" Global parameters \"\"\"\n","H = 512\n","W = 512\n","def create_dir(path):\n","    \"\"\" Create a directory. \"\"\"\n","    if not os.path.exists(path):\n","        os.makedirs(path)\n","\n","def load_data(path, split=0.2):\n","    images = sorted(glob(os.path.join(path,\"Image\", \"*.png\")))\n","    masks1 = sorted(glob(os.path.join(path,\"Left\", \"*.png\")))\n","    masks2 = sorted(glob(os.path.join(path,\"Right\", \"*.png\")))\n","\n","    split_size = int(len(images) * split)          # we are calculating the split size here\n","\n","    train_x, valid_x = train_test_split(images, test_size=split_size, random_state=42)  # training data and validation data will be selected randomly and the randomness seed is 42\n","    train_y1, valid_y1 = train_test_split(masks1, test_size=split_size, random_state=42)\n","    train_y2, valid_y2 = train_test_split(masks2, test_size=split_size, random_state=42)\n","\n","     \n","\n","    return (train_x, train_y1, train_y2), (valid_x, valid_y1, valid_y2) \n","\n","def read_image(path):\n","    x = cv2.imread(path, cv2.IMREAD_COLOR)\n","    x = cv2.resize(x, (W, H))\n","    x = x/255.0\n","    x = x.astype(np.float32)\n","    return x\n","\n","def read_mask(path1, path2):\n","    x1 = cv2.imread(path1, cv2.IMREAD_GRAYSCALE)\n","    x2 = cv2.imread(path2, cv2.IMREAD_GRAYSCALE)\n","    x = x1 + x2\n","    x = cv2.resize(x, (W, H))\n","    x = x/np.max(x)\n","    x = x > 0.5\n","    x = x.astype(np.float32)\n","    x = np.expand_dims(x, axis=-1)\n","    return x\n","\n","def tf_parse(x, y1, y2):\n","    def _parse(x, y1, y2):\n","        x = x.decode()\n","        y1 = y1.decode()\n","        y2 = y2.decode()\n","\n","        x = read_image(x)\n","        y = read_mask(y1, y2)\n","        return x, y\n","\n","    x, y = tf.numpy_function(_parse, [x, y1, y2], [tf.float32, tf.float32])\n","    x.set_shape([H, W, 3])\n","    y.set_shape([H, W, 1])\n","    return x, y\n","\n","def tf_dataset(X, Y1, Y2, batch=8):\n","    dataset = tf.data.Dataset.from_tensor_slices((X, Y1, Y2))\n","    dataset = dataset.shuffle(buffer_size=200)\n","    dataset = dataset.map(tf_parse)\n","    dataset = dataset.batch(batch)\n","    dataset = dataset.prefetch(4)\n","    return dataset\n","\n","if __name__ == \"__main__\":\n","    \"\"\" Seeding \"\"\"\n","    np.random.seed(42)\n","    tf.random.set_seed(42)\n","\n","    \"\"\" Directory for storing files \"\"\"\n","    create_dir(\"files\")\n","\n","    \"\"\" Hyperparameters \"\"\"\n","    batch_size = 8\n","    lr = 1e-5\n","    num_epochs = 50\n","    model_path = os.path.join(\"files\", \"Final_model_unet.h5\")\n","    csv_path = os.path.join(\"files\", \"Final_data_unet.csv\")\n","\n","    \"\"\" Dataset \"\"\"\n","    dataset_path = \"/content/drive/MyDrive/Colab\"\n","    (train_x, train_y1, train_y2), (valid_x, valid_y1, valid_y2)  = load_data(dataset_path)\n","\n","    print(f\"Train: {len(train_x)} - {len(train_y1)} - {len(train_y2)}\")\n","    print(f\"Valid: {len(valid_x)} - {len(valid_y1)} - {len(valid_y2)}\")\n","   \n","\n","    train_dataset = tf_dataset(train_x, train_y1, train_y2, batch=batch_size)\n","    valid_dataset = tf_dataset(valid_x, valid_y1, valid_y2, batch=batch_size)\n","\n","    \"\"\" Model \"\"\"\n","    model = build_unet((H, W, 3))\n","    metrics = [dice_coef, iou, Recall(), Precision()]\n","    #model.compile(loss=dice_loss, optimizer=Adam(lr), metrics=metrics)\n","    # model.compile(loss=dice_loss, optimizer=Adam(lr), metrics=[\"accuracy\"])\n","    model.compile(loss=dice_loss, optimizer=Adam(lr), metrics=[\"accuracy\",metrics])\n","\n","    callbacks = [\n","        ModelCheckpoint(model_path, verbose=1, save_best_only=True),\n","        ReduceLROnPlateau(monitor='val_loss', factor=0.1, patience=5, min_lr=1e-7, verbose=1),\n","        CSVLogger(csv_path)\n","    ]\n","\n","    model.fit(\n","        train_dataset,\n","        epochs=num_epochs,\n","        validation_data=valid_dataset,\n","        callbacks=callbacks\n","    )"],"metadata":{"id":"2eakjIUHucWW","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1664131209191,"user_tz":-360,"elapsed":18663991,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"c2178d05-2ad6-4d62-e89c-cba0425a385e"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Train: 1308 - 1308 - 1308\n","Valid: 326 - 326 - 326\n","Epoch 1/50\n","164/164 [==============================] - ETA: 0s - loss: 0.8224 - accuracy: 0.7763 - dice_coef: 0.1779 - iou: 0.0991 - recall: 0.9338 - precision: 0.1280\n","Epoch 1: val_loss improved from inf to 0.93046, saving model to files/model.h5\n","164/164 [==============================] - 1114s 7s/step - loss: 0.8224 - accuracy: 0.7763 - dice_coef: 0.1779 - iou: 0.0991 - recall: 0.9338 - precision: 0.1280 - val_loss: 0.9305 - val_accuracy: 0.1298 - val_dice_coef: 0.0695 - val_iou: 0.0360 - val_recall: 0.9929 - val_precision: 0.0382 - lr: 1.0000e-05\n","Epoch 2/50\n","164/164 [==============================] - ETA: 0s - loss: 0.6099 - accuracy: 0.9852 - dice_coef: 0.3901 - iou: 0.2433 - recall: 0.9523 - precision: 0.7162\n","Epoch 2: val_loss improved from 0.93046 to 0.87356, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.6099 - accuracy: 0.9852 - dice_coef: 0.3901 - iou: 0.2433 - recall: 0.9523 - precision: 0.7162 - val_loss: 0.8736 - val_accuracy: 0.9271 - val_dice_coef: 0.1263 - val_iou: 0.0675 - val_recall: 0.3590 - val_precision: 0.1982 - lr: 1.0000e-05\n","Epoch 3/50\n","164/164 [==============================] - ETA: 0s - loss: 0.5297 - accuracy: 0.9920 - dice_coef: 0.4704 - iou: 0.3079 - recall: 0.9614 - precision: 0.8347\n","Epoch 3: val_loss improved from 0.87356 to 0.60987, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.5297 - accuracy: 0.9920 - dice_coef: 0.4704 - iou: 0.3079 - recall: 0.9614 - precision: 0.8347 - val_loss: 0.6099 - val_accuracy: 0.9729 - val_dice_coef: 0.3900 - val_iou: 0.2425 - val_recall: 0.9024 - val_precision: 0.5700 - lr: 1.0000e-05\n","Epoch 4/50\n","164/164 [==============================] - ETA: 0s - loss: 0.5008 - accuracy: 0.9935 - dice_coef: 0.4992 - iou: 0.3329 - recall: 0.9661 - precision: 0.8633\n","Epoch 4: val_loss improved from 0.60987 to 0.49943, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.5008 - accuracy: 0.9935 - dice_coef: 0.4992 - iou: 0.3329 - recall: 0.9661 - precision: 0.8633 - val_loss: 0.4994 - val_accuracy: 0.9939 - val_dice_coef: 0.5004 - val_iou: 0.3340 - val_recall: 0.9677 - val_precision: 0.8707 - lr: 1.0000e-05\n","Epoch 5/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4798 - accuracy: 0.9945 - dice_coef: 0.5200 - iou: 0.3518 - recall: 0.9706 - precision: 0.8835\n","Epoch 5: val_loss improved from 0.49943 to 0.47520, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4798 - accuracy: 0.9945 - dice_coef: 0.5200 - iou: 0.3518 - recall: 0.9706 - precision: 0.8835 - val_loss: 0.4752 - val_accuracy: 0.9945 - val_dice_coef: 0.5249 - val_iou: 0.3562 - val_recall: 0.9765 - val_precision: 0.8786 - lr: 1.0000e-05\n","Epoch 6/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4636 - accuracy: 0.9948 - dice_coef: 0.5365 - iou: 0.3669 - recall: 0.9725 - precision: 0.8892\n","Epoch 6: val_loss improved from 0.47520 to 0.46740, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4636 - accuracy: 0.9948 - dice_coef: 0.5365 - iou: 0.3669 - recall: 0.9725 - precision: 0.8892 - val_loss: 0.4674 - val_accuracy: 0.9953 - val_dice_coef: 0.5328 - val_iou: 0.3634 - val_recall: 0.9304 - val_precision: 0.9343 - lr: 1.0000e-05\n","Epoch 7/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4511 - accuracy: 0.9949 - dice_coef: 0.5491 - iou: 0.3789 - recall: 0.9715 - precision: 0.8909\n","Epoch 7: val_loss improved from 0.46740 to 0.44628, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4511 - accuracy: 0.9949 - dice_coef: 0.5491 - iou: 0.3789 - recall: 0.9715 - precision: 0.8909 - val_loss: 0.4463 - val_accuracy: 0.9960 - val_dice_coef: 0.5539 - val_iou: 0.3834 - val_recall: 0.9637 - val_precision: 0.9254 - lr: 1.0000e-05\n","Epoch 8/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4343 - accuracy: 0.9957 - dice_coef: 0.5656 - iou: 0.3946 - recall: 0.9762 - precision: 0.9074\n","Epoch 8: val_loss improved from 0.44628 to 0.42826, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4343 - accuracy: 0.9957 - dice_coef: 0.5656 - iou: 0.3946 - recall: 0.9762 - precision: 0.9074 - val_loss: 0.4283 - val_accuracy: 0.9960 - val_dice_coef: 0.5717 - val_iou: 0.4006 - val_recall: 0.9691 - val_precision: 0.9205 - lr: 1.0000e-05\n","Epoch 9/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4212 - accuracy: 0.9960 - dice_coef: 0.5789 - iou: 0.4078 - recall: 0.9781 - precision: 0.9139\n","Epoch 9: val_loss improved from 0.42826 to 0.42002, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4212 - accuracy: 0.9960 - dice_coef: 0.5789 - iou: 0.4078 - recall: 0.9781 - precision: 0.9139 - val_loss: 0.4200 - val_accuracy: 0.9961 - val_dice_coef: 0.5798 - val_iou: 0.4086 - val_recall: 0.9645 - val_precision: 0.9255 - lr: 1.0000e-05\n","Epoch 10/50\n","164/164 [==============================] - ETA: 0s - loss: 0.4085 - accuracy: 0.9963 - dice_coef: 0.5917 - iou: 0.4205 - recall: 0.9790 - precision: 0.9192\n","Epoch 10: val_loss improved from 0.42002 to 0.40575, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.4085 - accuracy: 0.9963 - dice_coef: 0.5917 - iou: 0.4205 - recall: 0.9790 - precision: 0.9192 - val_loss: 0.4058 - val_accuracy: 0.9953 - val_dice_coef: 0.5944 - val_iou: 0.4232 - val_recall: 0.9784 - val_precision: 0.8963 - lr: 1.0000e-05\n","Epoch 11/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3984 - accuracy: 0.9962 - dice_coef: 0.6017 - iou: 0.4307 - recall: 0.9778 - precision: 0.9171\n","Epoch 11: val_loss improved from 0.40575 to 0.39662, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3984 - accuracy: 0.9962 - dice_coef: 0.6017 - iou: 0.4307 - recall: 0.9778 - precision: 0.9171 - val_loss: 0.3966 - val_accuracy: 0.9958 - val_dice_coef: 0.6035 - val_iou: 0.4324 - val_recall: 0.9729 - val_precision: 0.9133 - lr: 1.0000e-05\n","Epoch 12/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3849 - accuracy: 0.9966 - dice_coef: 0.6152 - iou: 0.4446 - recall: 0.9804 - precision: 0.9268\n","Epoch 12: val_loss improved from 0.39662 to 0.38407, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3849 - accuracy: 0.9966 - dice_coef: 0.6152 - iou: 0.4446 - recall: 0.9804 - precision: 0.9268 - val_loss: 0.3841 - val_accuracy: 0.9962 - val_dice_coef: 0.6158 - val_iou: 0.4452 - val_recall: 0.9683 - val_precision: 0.9263 - lr: 1.0000e-05\n","Epoch 13/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3733 - accuracy: 0.9968 - dice_coef: 0.6268 - iou: 0.4567 - recall: 0.9813 - precision: 0.9317\n","Epoch 13: val_loss improved from 0.38407 to 0.37555, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3733 - accuracy: 0.9968 - dice_coef: 0.6268 - iou: 0.4567 - recall: 0.9813 - precision: 0.9317 - val_loss: 0.3755 - val_accuracy: 0.9959 - val_dice_coef: 0.6246 - val_iou: 0.4544 - val_recall: 0.9719 - val_precision: 0.9154 - lr: 1.0000e-05\n","Epoch 14/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3626 - accuracy: 0.9969 - dice_coef: 0.6373 - iou: 0.4681 - recall: 0.9815 - precision: 0.9336\n","Epoch 14: val_loss improved from 0.37555 to 0.36678, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3626 - accuracy: 0.9969 - dice_coef: 0.6373 - iou: 0.4681 - recall: 0.9815 - precision: 0.9336 - val_loss: 0.3668 - val_accuracy: 0.9964 - val_dice_coef: 0.6333 - val_iou: 0.4637 - val_recall: 0.9558 - val_precision: 0.9425 - lr: 1.0000e-05\n","Epoch 15/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3512 - accuracy: 0.9972 - dice_coef: 0.6487 - iou: 0.4804 - recall: 0.9826 - precision: 0.9391\n","Epoch 15: val_loss improved from 0.36678 to 0.35680, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3512 - accuracy: 0.9972 - dice_coef: 0.6487 - iou: 0.4804 - recall: 0.9826 - precision: 0.9391 - val_loss: 0.3568 - val_accuracy: 0.9961 - val_dice_coef: 0.6432 - val_iou: 0.4744 - val_recall: 0.9595 - val_precision: 0.9307 - lr: 1.0000e-05\n","Epoch 16/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3403 - accuracy: 0.9973 - dice_coef: 0.6597 - iou: 0.4925 - recall: 0.9833 - precision: 0.9422\n","Epoch 16: val_loss improved from 0.35680 to 0.34732, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3403 - accuracy: 0.9973 - dice_coef: 0.6597 - iou: 0.4925 - recall: 0.9833 - precision: 0.9422 - val_loss: 0.3473 - val_accuracy: 0.9964 - val_dice_coef: 0.6527 - val_iou: 0.4848 - val_recall: 0.9551 - val_precision: 0.9414 - lr: 1.0000e-05\n","Epoch 17/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3301 - accuracy: 0.9974 - dice_coef: 0.6699 - iou: 0.5040 - recall: 0.9837 - precision: 0.9451\n","Epoch 17: val_loss improved from 0.34732 to 0.33674, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3301 - accuracy: 0.9974 - dice_coef: 0.6699 - iou: 0.5040 - recall: 0.9837 - precision: 0.9451 - val_loss: 0.3367 - val_accuracy: 0.9965 - val_dice_coef: 0.6632 - val_iou: 0.4965 - val_recall: 0.9527 - val_precision: 0.9464 - lr: 1.0000e-05\n","Epoch 18/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3200 - accuracy: 0.9976 - dice_coef: 0.6801 - iou: 0.5157 - recall: 0.9842 - precision: 0.9478\n","Epoch 18: val_loss improved from 0.33674 to 0.32814, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3200 - accuracy: 0.9976 - dice_coef: 0.6801 - iou: 0.5157 - recall: 0.9842 - precision: 0.9478 - val_loss: 0.3281 - val_accuracy: 0.9958 - val_dice_coef: 0.6720 - val_iou: 0.5063 - val_recall: 0.9757 - val_precision: 0.9104 - lr: 1.0000e-05\n","Epoch 19/50\n","164/164 [==============================] - ETA: 0s - loss: 0.3098 - accuracy: 0.9977 - dice_coef: 0.6902 - iou: 0.5274 - recall: 0.9850 - precision: 0.9515\n","Epoch 19: val_loss improved from 0.32814 to 0.32004, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.3098 - accuracy: 0.9977 - dice_coef: 0.6902 - iou: 0.5274 - recall: 0.9850 - precision: 0.9515 - val_loss: 0.3200 - val_accuracy: 0.9954 - val_dice_coef: 0.6799 - val_iou: 0.5154 - val_recall: 0.9793 - val_precision: 0.8972 - lr: 1.0000e-05\n","Epoch 20/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2998 - accuracy: 0.9979 - dice_coef: 0.7001 - iou: 0.5390 - recall: 0.9856 - precision: 0.9544\n","Epoch 20: val_loss improved from 0.32004 to 0.31439, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2998 - accuracy: 0.9979 - dice_coef: 0.7001 - iou: 0.5390 - recall: 0.9856 - precision: 0.9544 - val_loss: 0.3144 - val_accuracy: 0.9950 - val_dice_coef: 0.6858 - val_iou: 0.5221 - val_recall: 0.9828 - val_precision: 0.8871 - lr: 1.0000e-05\n","Epoch 21/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2899 - accuracy: 0.9980 - dice_coef: 0.7102 - iou: 0.5510 - recall: 0.9862 - precision: 0.9580\n","Epoch 21: val_loss improved from 0.31439 to 0.30247, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2899 - accuracy: 0.9980 - dice_coef: 0.7102 - iou: 0.5510 - recall: 0.9862 - precision: 0.9580 - val_loss: 0.3025 - val_accuracy: 0.9961 - val_dice_coef: 0.6978 - val_iou: 0.5362 - val_recall: 0.9697 - val_precision: 0.9213 - lr: 1.0000e-05\n","Epoch 22/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2805 - accuracy: 0.9981 - dice_coef: 0.7196 - iou: 0.5623 - recall: 0.9868 - precision: 0.9595\n","Epoch 22: val_loss improved from 0.30247 to 0.29397, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2805 - accuracy: 0.9981 - dice_coef: 0.7196 - iou: 0.5623 - recall: 0.9868 - precision: 0.9595 - val_loss: 0.2940 - val_accuracy: 0.9962 - val_dice_coef: 0.7058 - val_iou: 0.5456 - val_recall: 0.9612 - val_precision: 0.9323 - lr: 1.0000e-05\n","Epoch 23/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2713 - accuracy: 0.9982 - dice_coef: 0.7285 - iou: 0.5733 - recall: 0.9870 - precision: 0.9623\n","Epoch 23: val_loss improved from 0.29397 to 0.28641, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2713 - accuracy: 0.9982 - dice_coef: 0.7285 - iou: 0.5733 - recall: 0.9870 - precision: 0.9623 - val_loss: 0.2864 - val_accuracy: 0.9963 - val_dice_coef: 0.7136 - val_iou: 0.5550 - val_recall: 0.9669 - val_precision: 0.9287 - lr: 1.0000e-05\n","Epoch 24/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2629 - accuracy: 0.9983 - dice_coef: 0.7373 - iou: 0.5842 - recall: 0.9871 - precision: 0.9639\n","Epoch 24: val_loss improved from 0.28641 to 0.27769, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2629 - accuracy: 0.9983 - dice_coef: 0.7373 - iou: 0.5842 - recall: 0.9871 - precision: 0.9639 - val_loss: 0.2777 - val_accuracy: 0.9960 - val_dice_coef: 0.7222 - val_iou: 0.5654 - val_recall: 0.9711 - val_precision: 0.9180 - lr: 1.0000e-05\n","Epoch 25/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2541 - accuracy: 0.9984 - dice_coef: 0.7459 - iou: 0.5951 - recall: 0.9880 - precision: 0.9661\n","Epoch 25: val_loss improved from 0.27769 to 0.26906, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2541 - accuracy: 0.9984 - dice_coef: 0.7459 - iou: 0.5951 - recall: 0.9880 - precision: 0.9661 - val_loss: 0.2691 - val_accuracy: 0.9963 - val_dice_coef: 0.7307 - val_iou: 0.5758 - val_recall: 0.9637 - val_precision: 0.9323 - lr: 1.0000e-05\n","Epoch 26/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2451 - accuracy: 0.9985 - dice_coef: 0.7549 - iou: 0.6065 - recall: 0.9882 - precision: 0.9686\n","Epoch 26: val_loss improved from 0.26906 to 0.26337, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2451 - accuracy: 0.9985 - dice_coef: 0.7549 - iou: 0.6065 - recall: 0.9882 - precision: 0.9686 - val_loss: 0.2634 - val_accuracy: 0.9963 - val_dice_coef: 0.7369 - val_iou: 0.5838 - val_recall: 0.9597 - val_precision: 0.9370 - lr: 1.0000e-05\n","Epoch 27/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2369 - accuracy: 0.9986 - dice_coef: 0.7631 - iou: 0.6173 - recall: 0.9890 - precision: 0.9706\n","Epoch 27: val_loss improved from 0.26337 to 0.25802, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2369 - accuracy: 0.9986 - dice_coef: 0.7631 - iou: 0.6173 - recall: 0.9890 - precision: 0.9706 - val_loss: 0.2580 - val_accuracy: 0.9965 - val_dice_coef: 0.7420 - val_iou: 0.5901 - val_recall: 0.9495 - val_precision: 0.9500 - lr: 1.0000e-05\n","Epoch 28/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2294 - accuracy: 0.9986 - dice_coef: 0.7707 - iou: 0.6272 - recall: 0.9886 - precision: 0.9712\n","Epoch 28: val_loss improved from 0.25802 to 0.24781, saving model to files/model.h5\n","164/164 [==============================] - 338s 2s/step - loss: 0.2294 - accuracy: 0.9986 - dice_coef: 0.7707 - iou: 0.6272 - recall: 0.9886 - precision: 0.9712 - val_loss: 0.2478 - val_accuracy: 0.9965 - val_dice_coef: 0.7522 - val_iou: 0.6031 - val_recall: 0.9572 - val_precision: 0.9441 - lr: 1.0000e-05\n","Epoch 29/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2213 - accuracy: 0.9987 - dice_coef: 0.7787 - iou: 0.6378 - recall: 0.9892 - precision: 0.9730\n","Epoch 29: val_loss improved from 0.24781 to 0.24146, saving model to files/model.h5\n","164/164 [==============================] - 338s 2s/step - loss: 0.2213 - accuracy: 0.9987 - dice_coef: 0.7787 - iou: 0.6378 - recall: 0.9892 - precision: 0.9730 - val_loss: 0.2415 - val_accuracy: 0.9965 - val_dice_coef: 0.7586 - val_iou: 0.6113 - val_recall: 0.9560 - val_precision: 0.9443 - lr: 1.0000e-05\n","Epoch 30/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2140 - accuracy: 0.9987 - dice_coef: 0.7861 - iou: 0.6478 - recall: 0.9891 - precision: 0.9739\n","Epoch 30: val_loss improved from 0.24146 to 0.23540, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2140 - accuracy: 0.9987 - dice_coef: 0.7861 - iou: 0.6478 - recall: 0.9891 - precision: 0.9739 - val_loss: 0.2354 - val_accuracy: 0.9965 - val_dice_coef: 0.7645 - val_iou: 0.6191 - val_recall: 0.9569 - val_precision: 0.9430 - lr: 1.0000e-05\n","Epoch 31/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2071 - accuracy: 0.9987 - dice_coef: 0.7928 - iou: 0.6571 - recall: 0.9891 - precision: 0.9745\n","Epoch 31: val_loss improved from 0.23540 to 0.22935, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2071 - accuracy: 0.9987 - dice_coef: 0.7928 - iou: 0.6571 - recall: 0.9891 - precision: 0.9745 - val_loss: 0.2294 - val_accuracy: 0.9965 - val_dice_coef: 0.7706 - val_iou: 0.6272 - val_recall: 0.9479 - val_precision: 0.9507 - lr: 1.0000e-05\n","Epoch 32/50\n","164/164 [==============================] - ETA: 0s - loss: 0.2000 - accuracy: 0.9988 - dice_coef: 0.8000 - iou: 0.6669 - recall: 0.9890 - precision: 0.9757\n","Epoch 32: val_loss improved from 0.22935 to 0.22292, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.2000 - accuracy: 0.9988 - dice_coef: 0.8000 - iou: 0.6669 - recall: 0.9890 - precision: 0.9757 - val_loss: 0.2229 - val_accuracy: 0.9965 - val_dice_coef: 0.7772 - val_iou: 0.6359 - val_recall: 0.9524 - val_precision: 0.9471 - lr: 1.0000e-05\n","Epoch 33/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1927 - accuracy: 0.9988 - dice_coef: 0.8072 - iou: 0.6769 - recall: 0.9897 - precision: 0.9772\n","Epoch 33: val_loss improved from 0.22292 to 0.21544, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1927 - accuracy: 0.9988 - dice_coef: 0.8072 - iou: 0.6769 - recall: 0.9897 - precision: 0.9772 - val_loss: 0.2154 - val_accuracy: 0.9965 - val_dice_coef: 0.7846 - val_iou: 0.6457 - val_recall: 0.9589 - val_precision: 0.9415 - lr: 1.0000e-05\n","Epoch 34/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1864 - accuracy: 0.9989 - dice_coef: 0.8136 - iou: 0.6861 - recall: 0.9898 - precision: 0.9780\n","Epoch 34: val_loss improved from 0.21544 to 0.20932, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1864 - accuracy: 0.9989 - dice_coef: 0.8136 - iou: 0.6861 - recall: 0.9898 - precision: 0.9780 - val_loss: 0.2093 - val_accuracy: 0.9966 - val_dice_coef: 0.7907 - val_iou: 0.6540 - val_recall: 0.9510 - val_precision: 0.9513 - lr: 1.0000e-05\n","Epoch 35/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1796 - accuracy: 0.9989 - dice_coef: 0.8204 - iou: 0.6957 - recall: 0.9902 - precision: 0.9793\n","Epoch 35: val_loss improved from 0.20932 to 0.20382, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1796 - accuracy: 0.9989 - dice_coef: 0.8204 - iou: 0.6957 - recall: 0.9902 - precision: 0.9793 - val_loss: 0.2038 - val_accuracy: 0.9965 - val_dice_coef: 0.7961 - val_iou: 0.6617 - val_recall: 0.9567 - val_precision: 0.9433 - lr: 1.0000e-05\n","Epoch 36/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1736 - accuracy: 0.9989 - dice_coef: 0.8264 - iou: 0.7045 - recall: 0.9900 - precision: 0.9798\n","Epoch 36: val_loss improved from 0.20382 to 0.20197, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1736 - accuracy: 0.9989 - dice_coef: 0.8264 - iou: 0.7045 - recall: 0.9900 - precision: 0.9798 - val_loss: 0.2020 - val_accuracy: 0.9965 - val_dice_coef: 0.7980 - val_iou: 0.6641 - val_recall: 0.9380 - val_precision: 0.9606 - lr: 1.0000e-05\n","Epoch 37/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1676 - accuracy: 0.9990 - dice_coef: 0.8325 - iou: 0.7133 - recall: 0.9901 - precision: 0.9806\n","Epoch 37: val_loss improved from 0.20197 to 0.19248, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1676 - accuracy: 0.9990 - dice_coef: 0.8325 - iou: 0.7133 - recall: 0.9901 - precision: 0.9806 - val_loss: 0.1925 - val_accuracy: 0.9965 - val_dice_coef: 0.8075 - val_iou: 0.6774 - val_recall: 0.9529 - val_precision: 0.9482 - lr: 1.0000e-05\n","Epoch 38/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1619 - accuracy: 0.9990 - dice_coef: 0.8381 - iou: 0.7215 - recall: 0.9901 - precision: 0.9811\n","Epoch 38: val_loss improved from 0.19248 to 0.18989, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1619 - accuracy: 0.9990 - dice_coef: 0.8381 - iou: 0.7215 - recall: 0.9901 - precision: 0.9811 - val_loss: 0.1899 - val_accuracy: 0.9966 - val_dice_coef: 0.8101 - val_iou: 0.6812 - val_recall: 0.9415 - val_precision: 0.9590 - lr: 1.0000e-05\n","Epoch 39/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1559 - accuracy: 0.9990 - dice_coef: 0.8440 - iou: 0.7303 - recall: 0.9905 - precision: 0.9820\n","Epoch 39: val_loss improved from 0.18989 to 0.18290, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1559 - accuracy: 0.9990 - dice_coef: 0.8440 - iou: 0.7303 - recall: 0.9905 - precision: 0.9820 - val_loss: 0.1829 - val_accuracy: 0.9965 - val_dice_coef: 0.8171 - val_iou: 0.6910 - val_recall: 0.9574 - val_precision: 0.9418 - lr: 1.0000e-05\n","Epoch 40/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1511 - accuracy: 0.9990 - dice_coef: 0.8488 - iou: 0.7375 - recall: 0.9899 - precision: 0.9814\n","Epoch 40: val_loss improved from 0.18290 to 0.17813, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1511 - accuracy: 0.9990 - dice_coef: 0.8488 - iou: 0.7375 - recall: 0.9899 - precision: 0.9814 - val_loss: 0.1781 - val_accuracy: 0.9966 - val_dice_coef: 0.8219 - val_iou: 0.6978 - val_recall: 0.9460 - val_precision: 0.9567 - lr: 1.0000e-05\n","Epoch 41/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1455 - accuracy: 0.9991 - dice_coef: 0.8545 - iou: 0.7461 - recall: 0.9904 - precision: 0.9827\n","Epoch 41: val_loss improved from 0.17813 to 0.17236, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1455 - accuracy: 0.9991 - dice_coef: 0.8545 - iou: 0.7461 - recall: 0.9904 - precision: 0.9827 - val_loss: 0.1724 - val_accuracy: 0.9966 - val_dice_coef: 0.8277 - val_iou: 0.7062 - val_recall: 0.9525 - val_precision: 0.9504 - lr: 1.0000e-05\n","Epoch 42/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1404 - accuracy: 0.9991 - dice_coef: 0.8596 - iou: 0.7539 - recall: 0.9905 - precision: 0.9830\n","Epoch 42: val_loss improved from 0.17236 to 0.16796, saving model to files/model.h5\n","164/164 [==============================] - 338s 2s/step - loss: 0.1404 - accuracy: 0.9991 - dice_coef: 0.8596 - iou: 0.7539 - recall: 0.9905 - precision: 0.9830 - val_loss: 0.1680 - val_accuracy: 0.9966 - val_dice_coef: 0.8321 - val_iou: 0.7126 - val_recall: 0.9554 - val_precision: 0.9477 - lr: 1.0000e-05\n","Epoch 43/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1354 - accuracy: 0.9991 - dice_coef: 0.8646 - iou: 0.7616 - recall: 0.9906 - precision: 0.9838\n","Epoch 43: val_loss improved from 0.16796 to 0.16455, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1354 - accuracy: 0.9991 - dice_coef: 0.8646 - iou: 0.7616 - recall: 0.9906 - precision: 0.9838 - val_loss: 0.1645 - val_accuracy: 0.9966 - val_dice_coef: 0.8355 - val_iou: 0.7177 - val_recall: 0.9486 - val_precision: 0.9535 - lr: 1.0000e-05\n","Epoch 44/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1314 - accuracy: 0.9991 - dice_coef: 0.8686 - iou: 0.7679 - recall: 0.9900 - precision: 0.9832\n","Epoch 44: val_loss improved from 0.16455 to 0.15979, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1314 - accuracy: 0.9991 - dice_coef: 0.8686 - iou: 0.7679 - recall: 0.9900 - precision: 0.9832 - val_loss: 0.1598 - val_accuracy: 0.9966 - val_dice_coef: 0.8402 - val_iou: 0.7246 - val_recall: 0.9497 - val_precision: 0.9533 - lr: 1.0000e-05\n","Epoch 45/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1269 - accuracy: 0.9991 - dice_coef: 0.8731 - iou: 0.7749 - recall: 0.9899 - precision: 0.9835\n","Epoch 45: val_loss improved from 0.15979 to 0.15522, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1269 - accuracy: 0.9991 - dice_coef: 0.8731 - iou: 0.7749 - recall: 0.9899 - precision: 0.9835 - val_loss: 0.1552 - val_accuracy: 0.9966 - val_dice_coef: 0.8448 - val_iou: 0.7315 - val_recall: 0.9542 - val_precision: 0.9492 - lr: 1.0000e-05\n","Epoch 46/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1214 - accuracy: 0.9992 - dice_coef: 0.8786 - iou: 0.7836 - recall: 0.9910 - precision: 0.9852\n","Epoch 46: val_loss improved from 0.15522 to 0.15082, saving model to files/model.h5\n","164/164 [==============================] - 338s 2s/step - loss: 0.1214 - accuracy: 0.9992 - dice_coef: 0.8786 - iou: 0.7836 - recall: 0.9910 - precision: 0.9852 - val_loss: 0.1508 - val_accuracy: 0.9967 - val_dice_coef: 0.8493 - val_iou: 0.7383 - val_recall: 0.9492 - val_precision: 0.9545 - lr: 1.0000e-05\n","Epoch 47/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1171 - accuracy: 0.9992 - dice_coef: 0.8829 - iou: 0.7905 - recall: 0.9910 - precision: 0.9856\n","Epoch 47: val_loss improved from 0.15082 to 0.14712, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1171 - accuracy: 0.9992 - dice_coef: 0.8829 - iou: 0.7905 - recall: 0.9910 - precision: 0.9856 - val_loss: 0.1471 - val_accuracy: 0.9966 - val_dice_coef: 0.8528 - val_iou: 0.7436 - val_recall: 0.9482 - val_precision: 0.9546 - lr: 1.0000e-05\n","Epoch 48/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1133 - accuracy: 0.9992 - dice_coef: 0.8867 - iou: 0.7967 - recall: 0.9909 - precision: 0.9856\n","Epoch 48: val_loss improved from 0.14712 to 0.14381, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1133 - accuracy: 0.9992 - dice_coef: 0.8867 - iou: 0.7967 - recall: 0.9909 - precision: 0.9856 - val_loss: 0.1438 - val_accuracy: 0.9966 - val_dice_coef: 0.8562 - val_iou: 0.7489 - val_recall: 0.9537 - val_precision: 0.9493 - lr: 1.0000e-05\n","Epoch 49/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1093 - accuracy: 0.9992 - dice_coef: 0.8907 - iou: 0.8031 - recall: 0.9910 - precision: 0.9858\n","Epoch 49: val_loss improved from 0.14381 to 0.14023, saving model to files/model.h5\n","164/164 [==============================] - 339s 2s/step - loss: 0.1093 - accuracy: 0.9992 - dice_coef: 0.8907 - iou: 0.8031 - recall: 0.9910 - precision: 0.9858 - val_loss: 0.1402 - val_accuracy: 0.9967 - val_dice_coef: 0.8597 - val_iou: 0.7541 - val_recall: 0.9455 - val_precision: 0.9597 - lr: 1.0000e-05\n","Epoch 50/50\n","164/164 [==============================] - ETA: 0s - loss: 0.1057 - accuracy: 0.9992 - dice_coef: 0.8943 - iou: 0.8090 - recall: 0.9906 - precision: 0.9858\n","Epoch 50: val_loss improved from 0.14023 to 0.13571, saving model to files/model.h5\n","164/164 [==============================] - 338s 2s/step - loss: 0.1057 - accuracy: 0.9992 - dice_coef: 0.8943 - iou: 0.8090 - recall: 0.9906 - precision: 0.9858 - val_loss: 0.1357 - val_accuracy: 0.9966 - val_dice_coef: 0.8643 - val_iou: 0.7612 - val_recall: 0.9545 - val_precision: 0.9488 - lr: 1.0000e-05\n"]}]},{"cell_type":"code","source":["import os\n","os.environ[\"TF_CPP_MIN_LOG_LEVEL\"] = \"2\"\n","import numpy as np\n","import cv2\n","from glob import glob\n","from tqdm import tqdm\n","import tensorflow as tf\n","from tensorflow.keras.utils import CustomObjectScope\n"," \n","if __name__ == \"__main__\":\n"," \n","    \"\"\" Loading model \"\"\"\n","    with CustomObjectScope({'iou': iou, 'dice_coef': dice_coef, 'dice_loss': dice_loss}):\n","        model = tf.keras.models.load_model(\"files/Final_model_unet.h5\")\n","\n","    \"\"\" Dataset \"\"\"\n","    path = \"/content/drive/MyDrive/Test\"\n","    test=sorted(glob(os.path.join(path, \"*.png\")))\n","\n","    \"\"\" Predicting the mask \"\"\"\n","    for x in tqdm( test, total=len(test )):\n","        \"\"\" Extracing the image name. \"\"\"\n","        image_name = x.split(\"/\")[-1]\n","\n","        \"\"\" Reading the image \"\"\"\n","        img = cv2.imread(x, cv2.IMREAD_COLOR)\n","        img = cv2.resize(img, (W, H))\n","        x = img/255.0\n","        x = x.astype(np.float32)\n","        x = np.expand_dims(x, axis=0)\n","\n","         \n","\n","        \"\"\" Predicting the mask. \"\"\"\n","        y_pred = model.predict(x)[0] > 0.5\n","        y_pred = y_pred.astype(np.int32)\n","\n","        \"\"\" Saving the predicted mask along with the image and GT \"\"\"\n","        save_image_path = f\"/content/drive/MyDrive/Results_test/{image_name}\"\n","        y_pred = np.concatenate([y_pred, y_pred, y_pred], axis=-1)\n","\n","        sep_line = np.ones((H, 10, 3)) * 255\n","\n","        output_image = np.concatenate([img,sep_line, y_pred*255], axis=1)\n","        #cat_image = np.concatenate([sep_line,y_pred*255,sep_line], axis=1) # this line for just the output segment\n","        #cat_image = np.concatenate([ori_x,sep_line,y_pred*255], axis=1) # this line for just the output segment and the test image\n","        cv2.imwrite(save_image_path, output_image)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"VdA5n8Ad1P51","executionInfo":{"status":"ok","timestamp":1664131892517,"user_tz":-360,"elapsed":38238,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"62cb1787-25f6-4b45-e4a9-3765b5adee9f"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stderr","text":["100%|██████████| 145/145 [00:35<00:00,  4.05it/s]\n"]}]},{"cell_type":"markdown","source":["Outputs"],"metadata":{"id":"bnHJvEBxlrYc"}},{"cell_type":"code","source":["import cv2\n","import matplotlib.pyplot as plt\n","\n","x=cv2.imread(\"opt.png\")\n","x=cv2.cvtColor(x, cv2.COLOR_RGB2BGR)\n","plt.imshow(x)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":236},"id":"BDdeMYfjlqn9","executionInfo":{"status":"ok","timestamp":1665148899384,"user_tz":-360,"elapsed":2015,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"2c804d73-6ad7-42ee-a7bd-8af9619dbfbc"},"execution_count":2,"outputs":[{"output_type":"execute_result","data":{"text/plain":["<matplotlib.image.AxesImage at 0x7f568ec1fe90>"]},"metadata":{},"execution_count":2},{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["import pandas as pd\n","import matplotlib.pyplot as plt\n","data=pd.read_csv(\"Final_data_unet.csv\")\n"," "],"metadata":{"id":"SC1A6Fw5_O1i"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["print(data.head())"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"MB1pwIf9xwxZ","executionInfo":{"status":"ok","timestamp":1664161902873,"user_tz":-360,"elapsed":445,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"d07924ca-c8e8-458e-b5d0-540bc311db4f"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["   epoch  accuracy  dice_coef       iou      loss       lr  precision  \\\n","0      0  0.776284   0.177946  0.099066  0.822400  0.00001   0.127981   \n","1      1  0.985210   0.390065  0.243323  0.609936  0.00001   0.716223   \n","2      2  0.992031   0.470410  0.307866  0.529737  0.00001   0.834712   \n","3      3  0.993498   0.499227  0.332950  0.500817  0.00001   0.863348   \n","4      4  0.994522   0.520018  0.351776  0.479763  0.00001   0.883468   \n","\n","     recall  val_accuracy  val_dice_coef   val_iou  val_loss  val_precision  \\\n","0  0.933793      0.129814       0.069520  0.036018  0.930456       0.038240   \n","1  0.952332      0.927072       0.126261  0.067492  0.873564       0.198178   \n","2  0.961369      0.972882       0.390048  0.242541  0.609869       0.569959   \n","3  0.966058      0.993871       0.500406  0.333992  0.499427       0.870735   \n","4  0.970613      0.994480       0.524906  0.356233  0.475201       0.878577   \n","\n","   val_recall  \n","0    0.992950  \n","1    0.359001  \n","2    0.902425  \n","3    0.967715  \n","4    0.976501  \n"]}]},{"cell_type":"code","source":["train_loss=data.iloc[:,4:5].values\n"," "],"metadata":{"id":"sA2PHSFKx2s7"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["val_loss=data.iloc[:,11:12].values\n"," "],"metadata":{"id":"qvj0YwATyFLe"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_Acc=data.iloc[:,1:2].values\n"," "],"metadata":{"id":"A0evyf-PySrr"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["val_Acc=data.iloc[:,8:9].values\n"," "],"metadata":{"id":"II9Z0VQcyam3"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["train_iou=data.iloc[:,3:4].values\n"," "],"metadata":{"id":"8P_SB3CvyjaY"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["val_iou=data.iloc[:,10:11].values\n"," "],"metadata":{"id":"rF6TKKuDytX1"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["epoch=data.iloc[:,0:1].values"],"metadata":{"id":"HtuXaqUPy27q"},"execution_count":null,"outputs":[]},{"cell_type":"code","source":["print(f\"Accuracy = {train_Acc[-1]*100}%\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"7-2pWguEy8xW","executionInfo":{"status":"ok","timestamp":1664162591956,"user_tz":-360,"elapsed":458,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"f10a05a5-b6c0-4006-c2c7-c35566965480"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Accuracy = [99.91780519]%\n"]}]},{"cell_type":"code","source":["print(f\"IOU = {train_iou[-1]*100}%\")"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"IHgL5KRBzXH0","executionInfo":{"status":"ok","timestamp":1664162487852,"user_tz":-360,"elapsed":13,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"196e7e50-33c9-4fe1-ea0b-48d7912df827"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["IOU = [80.89572787]%\n"]}]},{"cell_type":"code","source":["plt.plot(epoch,train_loss,label='Train Loss')\n","plt.plot(epoch,val_loss,label='Validation Loss')\n","plt.legend()\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":265},"id":"2sV5uxfx0FhN","executionInfo":{"status":"ok","timestamp":1664162601377,"user_tz":-360,"elapsed":520,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"126fe913-2916-4654-8b58-473bc6dfa1ea"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["plt.plot(epoch,train_Acc,label='Train Accuracy')\n","plt.plot(epoch,val_Acc,label='Validation Accuracy')\n","plt.xlim(0, 5)\n","plt.ylim(0.8, 1)\n","plt.legend()\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":269},"id":"EwWfNqQh0hMI","executionInfo":{"status":"ok","timestamp":1664162743584,"user_tz":-360,"elapsed":586,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"78c7ceef-48bd-48e1-d0d5-a6ebeb5d1542"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":["plt.plot(epoch,train_iou,label='Train iou')\n","plt.plot(epoch,val_iou,label='Validation iou')\n","plt.legend()\n","plt.show()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/","height":265},"id":"xTA-0EZb0kQT","executionInfo":{"status":"ok","timestamp":1664162621220,"user_tz":-360,"elapsed":572,"user":{"displayName":"Shuvro Das","userId":"01513004482690500864"}},"outputId":"7938176a-4ea1-4467-8bcc-2e94628e8015"},"execution_count":null,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 432x288 with 1 Axes>"],"image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":[],"metadata":{"id":"NbR1gG990mCX"},"execution_count":null,"outputs":[]}]}