Python implementation for text generation using EEG signals from the brain
Thought2Text is a Python-based neural decoding pipeline for converting brain signals into text. It supports both invasive and non-invasive methods for speech decoding and typing reconstruction. The project includes features like multi-modality support, synthetic data generators, and privacy gating for intent classification.
- ▪Thought2Text is designed for both intracortical speech decoding and M/EEG-based typing reconstruction.
- ▪The pipeline includes preprocessing utilities and a CTC-based decoding method for handling variable-length sequences.
- ▪It allows for online, chunked decoding of neural streams and integrates with HuggingFace language models for improved accuracy.
Opening excerpt (first ~120 words) tap to expand
Thought2Text Thought2Text is a Python-based neural decoding pipeline designed for reproducing and experimenting with landmark brain-to-text systems. It supports both invasive intracortical speech decoding and non-invasive M/EEG-based typing reconstruction. Overview This project provides a modular framework for transforming neural signals into text. It implements a core workflow common to many state-of-the-art systems: neural signal -> preprocessing -> time-aligned neural features -> neural sequence model -> token probabilities -> beam search + language model -> text Key inspirations include: Willett et al. 2023 (Nature): High-performance speech neuroprosthesis using RNN phoneme decoders with CTC. Kunz et al. 2025 (Cell): Inner speech decoding with motor-intent gating and stack-gated RNNs.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.