Brain-Computer Interface (BCI) Project

• Background Information
• Core Features
• Limitations and Future Developments
• Environment Setup

Background Information

Main repo for the BCI division of Sheffield Bionics, a student-led society at the University of Sheffield. There are 3 main teams, namely Machine Learning, Signal Processing, and Data Collection, who are each responsible for a different step in the BCI-creation pipeline.

For 2024/25, we plan to develop a BCI that will allow individuals with motor impairments to play Atari Pong using only thoughts. Planning to implement an Attention-based LSTM model in PyTorch to detect thought patterns from an Ultracortex Mk. IV headset's EEG signals, supported by signal processing with SciPy and PostgreSQL for EEG data storage, with a target accuracy of 75-85%.

Members

Machine Learning team

Shohail Ismail

Signal Processing team

Thorpe Thorpe

Lucy Dalzell

Rowan Clarke

Nayantara Njondimackal

Adam Wieczorek

Lucy Dalzell

Data Collection team

Ethan Noulton

Core Features

Machine Learning team

TBD

Signal Processing team

TBD

Data Collection team

TBD

Limitations and Future Developments

TBD

Environment Setup

TBD