Data processing, model training, and testing

Check the code of the best team of the hackathon.

You will probably need some specific library to execute your model:

• Scipy: $ python -m pip install scipy
• Torch, torchvision, torchaudio (?) $ pip3 install torch torchvision torchaudio
• Cv_bridge

On windows, it is necessary to have ROS to obtain cv_bridge. Note, once you have a ROS, make sure your Python environment is connected to it.

Data preprocessing

The link of the code.

Modify the data path.

Model training

The link of the code.

Use the NEF server in Inria.

1. Request an account
2. Login
3. Run your code with some sample script

Model testing

You can test your model on your own PC with a bag playing. To do this you need to install ros on your computer first.