EEG2Text
About the project
Project description
The primary objective of the EEG2Text project is to create an innovative model capable of mapping non-invasive EEG signals directly into text. The team aims to develop a system that will allow for the reconstruction of thought content, based on the assumption that the processes of listening, reading, and thinking generate similar, decodable neural patterns.
EEG2Text is an advanced research project aimed at eliminating communication barriers by creating a system capable of turning thoughts into text. Currently, the team is focusing on transitioning from the phase of testing on ready-made datasets to building their own sample database. For this purpose, a proprietary application for collecting EEG data has been developed, and the first research sessions using a 32-channel BrainAccess cap are currently underway.
A key challenge of the project is to develop a multi-level model architecture that can handle the raw signal and transform it into a linguistic structure. In parallel, research is being conducted on cross-subject calibration - a method that will allow for the creation of a universal model despite individual differences in the brain activity of different users. In the future, the project aims to become a fully-fledged tool enabling interaction with a computer without the use of voice or movement.
Gallery
Project photos
Technologies