1) Clone the GIT repository with the "latest" version of the code:

git clone -b 2019 https://gitlab.com/mela64/motioninvariance

2) Setup a Python virtual environment (TensorFlow, OpenCV, and a few other packages are needed)

python3 -m venv python_env
source python_env/bin/activate
pip3 install tensorflow
pip3 install opencv-python
pip3 install scipy

3) Get the list of parameters

python vprocessor.py

4) Rub a basic experiment

python vprocessor.py --run videos/skater.avi --out test_output --port 7123 --gray 1 --save_scores_only 0 --res 240x110 --rep 10 --m 3 --f 7 

5) Open your web browser, and browse this address: 

http://127.0.0.1:7123

5) You can also monitor the evolution of the system using Tensorboard:

tensorboard --logdir <exp_path>/model/tensorboard

where <exp_path> is the output folder of the experiment, that, in this case, is: test_output

It will open a server on port 6006:

http://127.0.0.1:6006