CIPHER: Conformer-based Inference of Phonemes from High-density EEG
For researchers in EEG-based speech decoding, this paper provides a benchmark and confound analysis, but the results are incremental and highlight current limitations.
CIPHER uses a dual-pathway conformer model to decode phonemes from high-density EEG, achieving near-ceiling performance on binary tasks but limited fine-grained discriminability on an 11-class CVC task (WER ~0.67-0.69). The work is positioned as a benchmark and feature-comparison study rather than a practical EEG-to-text system.
Decoding speech information from scalp EEG remains difficult due to low SNR and spatial blurring. We present CIPHER (Conformer-based Inference of Phonemes from High-density EEG Representations), a dual-pathway model using (i) ERP features and (ii) broadband DDA coefficients. On OpenNeuro ds006104 (24 participants, two studies with concurrent TMS), binary articulatory tasks reach near-ceiling performance but are highly confound-vulnerable (acoustic onset separability and TMS-target blocking). On the primary 11-class CVC phoneme task under full Study 2 LOSO (16 held-out subjects), performance is substantially lower (real-word WER: ERP 0.671 +/- 0.080, DDA 0.688 +/- 0.096, indicating limited fine-grained discriminability. We therefore position this work as a benchmark and feature-comparison study rather than an EEG-to-text system, and we constrain neural-representation claims to confound-controlled evidence.