Electric Sheep

Inspiration

We were driven by the challenge to make communication accessible for all, especially for those with mobility limitations. The idea was to create a tool allowing users to convey their thoughts effortlessly and without physical constraints.

What it does

CortexType facilitates typing through mind control, harnessing brain-computer interface technology to translate thoughts into text on a computer.

How we built it

We integrated EEG technology with advanced algorithms to interpret brain signals, particularly the P300 response, creating a thought-to-text conversion process.

Challenges we ran into

Balancing accuracy and response time was a significant challenge.

Accomplishments that we're proud of

We are proud of testing out different models to improve the prediction performance.

What we learned

• It takes time to set up all software and hardware correctly, so we need to start earlier next time.
• The EEG signal is very noisy and we might need to consider some cool denoising techniques.
  

What's next for CortexType

• Experiment with other EEG signal preprocessing techniques.
• Complete testing and adaptation of models developed during the hackathon using the EEG equipment.
• Consider introducing an additional subject-specific layer to account for variability between users.
  

Built With

• oculus
• python
• unity