fabriciotorquato / pyxavier

This work evaluates different recurrent neural network architectures to control a virtual object on Robot Operating System (ROS) using electroencephalogram for signal acquisition. For the interface controls, voluntary hand motor actions were used, each hand for a different direction.

Pyxavier

Requirements

• Git
• Virtualvenv

Library

Python

Description	Tech	Resources
get raw data - EEG	Emokit	API

Setup project

./setup.sh

Run

source venv/bin/activate

python -m example.create_dataset --dir={bci/emotion/ihc} --full={True/False}

python -m example.training_model --dir={bci/emotion/ihc} --show={True/False} --full={True/False}

python -m example.run_mlp --dir={bci/emotion/ihc} --show={True/False} --full={True/False}

python -m example.run_cnn --dir={bci/emotion/ihc} --show={True/False} --full={True/False}

python -m example.run_rnn --dir={bci/emotion/ihc} --show={True/False} --full={True/False}