Mask_RCNN

Mask_RCNN copied to clipboard

rizwan78910

Hi everyone; would it be possible to use the model to do material detection (between wood, plastic, metal,etc) instead of object detection? If yes, how can it be done? Best...

sguarnerAlten

my dataset was annotated by labelme image annotator, but i tried convert vgg images annotator json format. anyone can help me to fix an correct json format

amuthalingeswaranbose

--------------------------------------------------------------------------- OSError Traceback (most recent call last) /var/folders/85/ct6fl3cd7jb4xqryxdmdp24m0000gp/T/ipykernel_41965/1575341098.py in 18 model.load_weights(COCO_WEIGHTS_PATH, by_name=True, exclude=["mrcnn_class_logits", "mrcnn_bbox_fc", "mrcnn_bbox", "mrcnn_mask"]) 19 # train weights (output layers or 'heads') ---> 20 model.train(dataset_train, dataset_train, learning_rate=config.LEARNING_RATE, epochs=1,...

ftrajano

Polygon coordinates for the mask of each detected object

6

Hi i was wondering that can we get the coordinates of the polygon mask generated of the each detected object using Mask_RCNN.? if so how can we get that.?

iresusharma

"False value" in x,y coodinates of mask

2

In the predict value of mask im gettin "False value" and not the coodinates,why? This is the code:  print("Images: {}\nClasses: {}".format(len(dataset.image_ids), dataset.class_names)) path_to_new_image = '/content/drive/MyDrive/MaskRCNN/Stomata/train/ 73-ab1.jpg' image1 = mpimg.imread(path_to_new_image)...

claudio9russo7

Map per class on custom dataset

10

Is it possible to produce the map per class on a custom dataset?

tomklaver1995

Trying to access the location or coordinates of predicted masks Mask-RCNN

6

We are working on a project, vision controlled robot for pick and place partial occluded objects. We are using mask RCNN architecture. Until now, we are successful in producing masks,...

DanishMalik333

Made Changes in model.py, utils.py and requirements.txt

ParthGaneriwala

Occurs with the new DataGenerator with Sequence ``` File ~/VSCode/ai1/mrcnn/model.py:2590, in MaskRCNN.train(self, train_dataset, val_dataset, learning_rate, epochs, layers, augmentation, custom_callbacks, no_augmentation_sources) 2587 else: 2588 workers = multiprocessing.cpu_count() -> 2590 self.keras_model.fit( 2591...

trueToastedCode