Synthesizing Speech from Intracranial Depth Electrodes using an Encoder-Decoder Framework

Abstract

Speech Neuroprostheses have the potential to enable communication for people with dysarthria or anarthria. Recent advances have demonstrated high-quality text decoding and speech synthesis from electrocorticographic grids placed on the cortical surface. Here, we investigate a less invasive measurement modality in three participants, namely stereotactic EEG (sEEG) that provides sparse sampling from multiple brain regions, including subcortical regions. To evaluate whether sEEG can also be used to synthesize audio from neural recordings, we employ a recurrent encoder-decoder model based on modern deep learning methods. We find that speech can indeed be reconstructed with correlations up to 0.8 from these minimally invasive recordings, despite limited amounts of training data. In particular, the architecture we employ naturally picks up on the temporal nature of the data and thereby outperforms an existing benchmark based on non-regressive convolutional neural networks.

References in corpus (5)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Neural Machine Translation by Jointly Learning to Align and Translate
• Decoupled Weight Decay Regularization
• Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions
• Language Models as Few-Shot Learner for Task-Oriented Dialogue Systems

Cited by in corpus (1)

• Brain-to-Speech: Prosody Feature Engineering and Transformer-Based Reconstruction