Mask R-CNN is a deep learning model for computer vision developed by the Facebook AI group that achieves state-of-the-art results on semantic segmentation (object recognition and pixel labeling) tasks. An implementation of the model is made available by Matterport on their github page. The code in their repo works with MS Coco (a benchmark dataset for semantic segmentation) out of the box, but provides for easy extensibility to any kind of dataset or image segmentation task.
This is a fork of the matterport/mask_rcnn repo that we have set up to integrate with Weights and Biases (wandb). wandb is a cloud interface for tracking model parameters and performance, allowing machine learning teams to coordinate work in a way similar to github. Here are the results of our runs. For a more detailed overview of our process and results see our blog post.
We have also streamlined the setup process of the original repo to get it up and running quickly on the tensorflow_p36 environment of the AWS Deep Learning AMI (Ubuntu) Version 10.0. To do so, start up an instance with at least 100 GB of storage, ssh into it, and do:
source activate tensorflow_p36
git clone https://github.com/connorhough/mask_rcnn
pip install cython
pip install -r requirements.txt
pip install tensorflow-gpu==1.7.0
python setup.py install
wandb init, then follow the init steps
wandb run samples/coco/coco.py train --model=imagenet --dataset=../coco --download=True
After the first run, use the above command without the
The parameter sweep can be run with