SYSTEM AND METHOD FOR SILENT SPEECH DECODING

Systems and methods are provided for decoding the silent speech of a user. Speech of the user (e.g., silent, vocalized, etc.) may be detected and captured by a speech input device configured to measure signals indicative of the speech muscle activation patterns of the user. A trained machine learning model configured to decode the speech of the user based at least in part on the signal indicative of the speech muscle activation patterns of the user. To improve accuracy of the model, the trained machine learning model may be trained using training data obtained in at least a subset of sampling contexts of a plurality of sampling contexts. At least one processor may be configured to output the decoded speech of the user.