This is a fork of Matterport's Mask R-CNN implementation modified to function with python2. It has not yet been tested to see if the model still trains or not, and I've definitely broken some functionality (downloading pretrained models, etc.), but if you just need image segmentation with a coco-pretrained model in python2, you're in luck.
-
Clone the repository.
-
Install dependencies
pip install -r requirements.txt
-
Optional: Run setup from the repository root directory
python setup.py install
-
Download pre-trained COCO weights (mask_rcnn_coco.h5) from the releases page, place it in the root directory of this package.
-
Execute
python pretrained_test.py
with a python2 interpreter. This program uses PretrainedLoader to construct and load a model, then evaluates it on the images provided with the package.
To include this in your project, import PretrainedLoader
as in pretrained_test.py.
Happy segmenting!