Skip to content

Code of my Bachelor's thesis on Brain-Computer-Interfaces, Deep Transfer Learning with CNNs, Hyperdimensional Computing and EEG data processing.

Notifications You must be signed in to change notification settings

danielbinschmid/portable-bci

Repository files navigation

portable-bci

This repository contains all the code written and used during a bachelor's thesis about Usability driven design of learning algorithms on a wearable consumer-grade brain-computer interface. Three different programming languages are used, javascript, C++, and python. Experimental comparisons are conducted, and a javascript based Smartwatch application is implemented. Main programming language is javascript; knowledge in C++ is helpful to understand the full digitial signal processing pipeline; the ecosystem of python is used to accelerate experiments.

Author: Daniel Bin Schmid
Degree: B.Sc. Medieninformatik
Mail: [email protected]
Chair & supervisor: ITI/ STAR at Universität Stuttgart; Jun.- Prof. Hussam Amrouch

Content

Content is structured in four subfolders.

  • cpp_env - C++ code. Digital signal processing with iir1, and Riemannian Geometry of covariance matrices with RigByCpp

  • node_env - NodeJs experiments with Node shell. Uses CUDA GPU to accelerate experiments. TensorflowJs is used to facilitate machine learning experiments.

  • py_env - Machine learning experiments with python

  • watch_webapp - The Smartwatch application, written in javascript. Implemented with Vue, Vuetify, and Cordova.

Submodules

Remember to init submodules by

git submodule update --init --recursive

Data

To download the datasets and move them into their according folders run the bash script download_data.sh

About

Code of my Bachelor's thesis on Brain-Computer-Interfaces, Deep Transfer Learning with CNNs, Hyperdimensional Computing and EEG data processing.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published