This is an implementation of Mask R-CNN on Python 3, Keras, and TensorFlow. The model generates bounding boxes and segmentation masks for each instance of an object in the image. It's based on Feature Pyramid Network (FPN) and a ResNet101 backbone.

• Tensorflow 1.14
• Keras 2.1.0

$ pip install -r requirements.txt

If resolver has problems :

$ pip install -r requirements.txt --use-feature=2020-resolver

Put the last weights in the logs directory : painting20200830T0323.
Then you are ready to go.

You will find the weights here : https://drive.google.com/file/d/1UEP8uPDlCuFGiWvvyziuEH7akmNfGgZX/view?usp=sharing

$ cd data
$ python painting.py train --dataset=dataset --weights=last

• If memory is exceeding reduce the number of images al line 66 in painting.py
• The model is compatible with tf 1.14 not 2