Changes in model.py and annotations. added attribute annotations

datasets/feather_damage/train/coco_annotations.json

@@ -487065,10 +487065,10 @@
  ],
  "categories": [
   {
-   "id": 0,
+   "id": 1,
    "name": "Hen",
    "supercategory": "root"
   }
  ],
  "tag_categories": []
-}
+}

datasets/feather_damage/val/coco_annotations.json

@@ -487065,10 +487065,10 @@
  ],
  "categories": [
   {
-   "id": 0,
+   "id": 1,
    "name": "Hen",
    "supercategory": "root"
   }
  ],
  "tag_categories": []
-}
+}

mrcnn/model.py

@@ -1240,7 +1240,6 @@ def load_image_gt(dataset, config, image_id, augmentation=None):
         max_dim=config.IMAGE_MAX_DIM,
         mode=config.IMAGE_RESIZE_MODE)
     mask = utils.resize_mask(mask, scale, padding, crop)
-    print(image.shape)
 
     # Augmentation
     # This requires the imgaug lib (https://github.com/aleju/imgaug)
@@ -1683,7 +1682,7 @@ def __init__(self, dataset, config, shuffle=True, augmentation=None,
                  random_rois=0, detection_targets=False):
         print("Datagenerator called") 
         self.image_ids = np.copy(dataset.image_ids)
-        print(self.image_ids)
+        #print(self.image_ids)
         self.dataset = dataset
         self.config = config
 
@@ -1703,8 +1702,6 @@ def __init__(self, dataset, config, shuffle=True, augmentation=None,
         self.detection_targets = detection_targets
 
     def __len__(self):
-        print("Length:")
-        print(int(np.ceil(len(self.image_ids) / float(self.batch_size))))
         return int(np.ceil(len(self.image_ids) / float(self.batch_size)))
 
     def __getitem__(self, idx):

mrcnn/utils.py

@@ -382,6 +382,7 @@ def load_mask(self, image_id):
         logging.warning("You are using the default load_mask(), maybe you need to define your own one.")
         mask = np.empty([0, 0, 0])
         class_ids = np.empty([0], np.int32)
+        print(mask)
         return mask, class_ids
 
 

requirements.txt

@@ -8,4 +8,4 @@ tensorflow>=2.0.0
 opencv-python
 h5py
 imgaug
-IPython[all]
+IPython[all]

samples/feather_damage/training.py

@@ -0,0 +1,232 @@
+import os
+import sys
+import json
+import numpy as np
+import time
+from PIL import Image, ImageDraw
+
+# Set the ROOT_DIR variable to the root directory of the Mask_RCNN git repo
+ROOT_DIR = '../../'
+assert os.path.exists(ROOT_DIR), 'ROOT_DIR does not exist. Did you forget to read the instructions above? ;)'
+
+# Import mrcnn libraries
+sys.path.append(ROOT_DIR) 
+from mrcnn.config import Config
+import mrcnn.utils as utils
+from mrcnn import visualize
+import mrcnn.model as modellib
+
+# Directory to save logs and trained model
+MODEL_DIR = "/media/christian/SamsungSSD/tensorflow_logs/"
+
+# Local path to trained weights file
+COCO_MODEL_PATH = os.path.join(ROOT_DIR, "mask_rcnn_coco.h5")
+
+# Download COCO trained weights from Releases if needed
+if not os.path.exists(COCO_MODEL_PATH):
+    utils.download_trained_weights(COCO_MODEL_PATH)
+
+class CigButtsConfig(Config):
+    """Configuration for training on the cigarette butts dataset.
+    Derives from the base Config class and overrides values specific
+    to the cigarette butts dataset.
+    """
+    # Give the configuration a recognizable name
+    NAME = "feather__damage"
+
+    # Train on 1 GPU and 1 image per GPU. Batch size is 1 (GPUs * images/GPU).
+    GPU_COUNT = 1
+    IMAGES_PER_GPU = 1
+
+    # Number of classes (including background)
+    NUM_CLASSES = 1 + 2  # background + 2 (feather damage + no feather damage)
+
+    # All of our training images are 512x512
+    IMAGE_MIN_DIM = None
+    IMAGE_MAX_DIM = 512 # must be dividable by 2 at least 6 times 
+
+    # You can experiment with this number to see if it improves training
+    STEPS_PER_EPOCH = 50 # 500
+
+    # This is how often validation is run. If you are using too much hard drive space
+    # on saved models (in the MODEL_DIR), try making this value larger.
+    VALIDATION_STEPS = 10 # 5
+
+    # Matterport originally used resnet101, but I downsized to fit it on my graphics card
+    BACKBONE = 'resnet50'
+
+    # To be honest, I haven't taken the time to figure out what these do
+    RPN_ANCHOR_SCALES = (8, 16, 32, 64, 128)
+    TRAIN_ROIS_PER_IMAGE = 32
+    MAX_GT_INSTANCES = 20 
+    POST_NMS_ROIS_INFERENCE = 500 
+    POST_NMS_ROIS_TRAINING = 1000 
+
+config = CigButtsConfig()
+config.display()
+
+class CocoLikeDataset(utils.Dataset):
+    """ Generates a COCO-like dataset, i.e. an image dataset annotated in the style of the COCO dataset.
+        See http://cocodataset.org/#home for more information.
+    """
+    def load_data(self, annotation_json, images_dir):
+        """ Load the coco-like dataset from json
+        Args:
+            annotation_json: The path to the coco annotations json file
+            images_dir: The directory holding the images referred to by the json file
+        """
+        # Load json from file
+        json_file = open(annotation_json)
+        coco_json = json.load(json_file)
+        json_file.close()
+
+        # Add the class names using the base method from utils.Dataset
+        source_name = "coco_like"
+        for category in coco_json['categories']:
+            class_id = category['id'] 
+            class_name = category['name'] 
+            if class_id < 1:
+                print('Error: Class id for "{}" cannot be less than one. (0 is reserved for the background)'.format(class_name))
+                return
+
+            self.add_class(source_name, class_id, class_name)
+
+        # Get all annotations
+        annotations = {}
+        for annotation in coco_json['annotations']:
+            image_id = annotation['image_id']
+            if image_id not in annotations:
+                annotations[image_id] = []
+            annotations[image_id].append(annotation)
+
+        # Get all images and add them to the dataset
+        seen_images = {}
+        for image in coco_json['images']:
+            image_id = image['id']
+            if image_id in seen_images:
+                print("Warning: Skipping duplicate image id: {}".format(image))
+            else:
+                seen_images[image_id] = image
+                try:
+                    #image_file_name = format(image['id'], '08d')+"_"+image['file_name']
+                    image_file_name = image['file_name'] 
+                    image_width = image['width']
+                    image_height = image['height']
+                except KeyError as key:
+                    print("Warning: Skipping image (id: {}) with missing key: {}".format(image_id, key))
+
+                image_path = os.path.abspath(os.path.join(images_dir, image_file_name))
+                image_annotations = annotations[image_id]
+                hasCrowd = False # Quick fix to avoid problems with "count" annotations. If more than 1segmentation/object -> skip image
+                for element in image_annotations:
+                  if element['iscrowd']:
+                    hasCrowd = True 
+
+                # Add the image using the base method from utils.Dataset (if file exists)
+                if os.path.isfile(image_path):
+                 if not hasCrowd: 
+                  self.add_image(
+                      source=source_name,
+                      image_id=image_id,
+                      path=image_path,
+                      width=image_width,
+                      height=image_height,
+                      annotations=image_annotations
+                  )
+
+
+    def load_mask(self, image_id):
+        """ Load instance masks for the given image.
+        MaskRCNN expects masks in the form of a bitmap [height, width, instances].
+        Args:
+            image_id: The id of the image to load masks for
+        Returns:
+            masks: A bool array of shape [height, width, instance count] with
+                one mask per instance.
+            class_ids: a 1D array of class IDs of the instance masks.
+        """
+        image_info = self.image_info[image_id]
+        annotations = image_info['annotations']
+        instance_masks = []
+        class_ids = []
+
+        for annotation in annotations:
+            class_id = annotation['category_id']
+            mask = Image.new('1', (image_info['width'], image_info['height']))
+            mask_draw = ImageDraw.ImageDraw(mask, '1')
+            for segmentation in annotation['segmentation']:
+                mask_draw.polygon(segmentation, fill=1)
+                bool_array = np.array(mask) > 0
+                instance_masks.append(bool_array)
+                class_ids.append(class_id)
+
+        mask = np.dstack(instance_masks)
+        class_ids = np.array(class_ids, dtype=np.int32)
+
+        return mask, class_ids
+
+dataset_train = CocoLikeDataset()
+dataset_train.load_data('/media/christian/SamsungSSD/ZED/datasets/900_images/attribute_annotations.json', '/media/christian/SamsungSSD/ZED/datasets/900_images/train/')
+dataset_train.prepare()
+
+dataset_val = CocoLikeDataset()
+dataset_val.load_data('/media/christian/SamsungSSD/ZED/datasets/900_images/attribute_annotations.json', '/media/christian/SamsungSSD/ZED/datasets/900_images/val/')
+dataset_val.prepare()
+
+print(dataset_train)
+#dataset = dataset_train
+#image_ids = np.random.choice(dataset.image_ids, 4)
+#for image_id in image_ids:
+#    image = dataset.load_image(image_id)
+#    #print(image.shape)
+#    mask, class_ids = dataset.load_mask(image_id)
+#    visualize.display_top_masks(image, mask, class_ids, dataset.class_names)
+
+#### Training ####
+
+# Create model in training mode
+model = modellib.MaskRCNN(mode="training", config=config,
+                          model_dir=MODEL_DIR)
+
+# Which weights to start with?
+init_with = "coco"  # imagenet, coco, or last
+
+if init_with == "imagenet":
+    model.load_weights(model.get_imagenet_weights(), by_name=True)
+elif init_with == "coco":
+    # Load weights trained on MS COCO, but skip layers that
+    # are different due to the different number of classes
+    # See README for instructions to download the COCO weights
+    model.load_weights(COCO_MODEL_PATH, by_name=True,
+                       exclude=["mrcnn_class_logits", "mrcnn_bbox_fc", 
+                                "mrcnn_bbox", "mrcnn_mask"])
+elif init_with == "last":
+    # Load the last model you trained and continue training
+    model.load_weights(model.find_last(), by_name=True)
+
+# Train the head branches
+# Passing layers="heads" freezes all layers except the head
+# layers. You can also pass a regular expression to select
+# which layers to train by name pattern.
+
+#start_train = time.time()
+#model.train(dataset_train, dataset_val, 
+#            learning_rate=config.LEARNING_RATE, 
+#            epochs=10, 
+#            layers='heads')
+#end_train = time.time()
+#minutes = round((end_train - start_train) / 60, 2)
+#print(f'Training took {minutes} minutes')
+
+# Fine tune all layers
+# Passing layers="all" trains all layers. You can also 
+# pass a regular expression to select which layers to
+# train by name pattern.
+start_train = time.time()
+model.train(dataset_train, dataset_val, 
+            learning_rate=config.LEARNING_RATE / 10,
+            epochs=100, 
+            layers="all")
+end_train = time.time()
+minutes = round((end_train - start_train) / 60, 2)
+print(f'Training took {minutes} minutes')