Install docker and nvidia-docker

Clone the repository:

git clone https://github.com/azati/nasa2017

Build docker image:

cd nasa2017/backend/docker_image
docker build -t theano_gpu .

Unpack roofs images:

cd ../roofs_classifying/data
unrar x -r all_buildings.part1.rar

Download pre-trained weights for ResNet-50:

cd ..
wget https://s3.amazonaws.com/lasagne/recipes/pretrained/imagenet/resnet50.pkl

Run jupyter:

./run_jupyter.sh

Navigate http://localhost:8888 and open file all_buildings_prediction-resnet.ipynb

Run all the cells in the notebook to perform roofs classification.

Install CocoaPods

Run pod install from Terminal, then open your app's .xcworkspace file to launch Xcode.

https://azati.com