Komeiji, S. | Search bioRxiv

bioRxiv2024

Speech Synthesis from Electrocorticogram During Imagined Speech Using a Transformer-Based Decoder and Pretrained Vocoder

Synthesizing speech from Electrocorticogram (ECoG) signals recorded during imagined speech remains a challenge due to the absence of synchronized audio signals for training. To address this, we propose a training framework that utilizes audio recorded during overt speech tasks as a surrogate ground truth for imagined speech signals, based on the consistency of the linguistic content. We employed a Transformer-based decoder to generate log-mel spectrograms from imagined speech ECoG, which were then converted into waveform audio using a pre-trained Parallel WaveGAN. In experiments involving ECoG recordings from 13 participants, the synthesized speech achieved dynamic time warping-aligned Pearson correlation coefficients ranging from 0.74 to 0.84 with the proxy targets. These results demonstrate that overt speech audio can serve as an effective training target for reconstructing imagined speech, offering a viable solution for training decoders in the absence of behavioral output.

neuroscience↗

bioRxiv2024

Feasibility of decoding covert speech in ECoG with aTransformer trained on overt speech

Several attempts for speech brain-computer interfacing (BCI) have been made to decode phonemes, sub-words, words, or sentences using invasive measurements, such as the electrocorticogram (ECoG), during auditory speech perception, overt speech, or imagined (covert) speech. Decoding sentences from covert speech is a challenging task. Sixteen epilepsy patients with intracranially implanted electrodes participated in this study, and ECoGs were recorded during overt speech and covert speech of eight Japanese sentences, each consisting of three tokens. In particular, Transformer neural network model was applied to decode text sentences from covert speech, which was trained using ECoGs obtained during overt speech. We first examined the proposed Transformer model using the same task for training and testing, and then evaluated the models performance when trained with overt task for decoding covert speech. The Transformer model trained on covert speech achieved an average token error rate (TER) of 46.6% for decoding covert speech, whereas the model trained on overt speech achieved a TER of 46.3% (p > 0.05; d = 0.07). Therefore, the challenge of collecting training data for covert speech can be addressed using overt speech. The performance of covert speech can improve by employing several overt speeches.

neuroscience↗