Researchers have developed PhonoBridge, a novel two-stage framework for reconstructing silent speech from Electromyography (EMG) signals into text. The system utilizes a channel-aware EMG-to-phoneme frontend and a large language model (LLM) backend. By using phonemes as an intermediate representation, PhonoBridge expands its training data by converting existing text transcripts into phoneme-text pairs, and employs an Error-Profile-Guided Augmentation technique to improve the LLM's robustness to frontend noise. This approach achieved a 24.77% Word Error Rate (WER) on the Open Vocabulary silent-speech dataset, surpassing previous benchmarks. AI
IMPACT This research advances silent speech recognition, potentially enabling new human-computer interaction methods.
RANK_REASON The cluster contains an academic paper detailing a new method for speech reconstruction. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
