Show HN: Subvocal – Talk to your computer without making noise
The Subvocal Company has introduced smart glasses that use radio frequency (RF) signals to decode silent mouth movements and internal speech into text in real time. Initially experimenting with facial electromyography (EMG) electrodes, the founders transitioned to an integrated RF sensing system embedded within glasses frames. This technology enables discreet, hands-free text input and computing in public spaces or quiet environments where speaking aloud is socially awkward or prohibited. It represents a significant step forward in human-computer interaction (HCI) and accessibility for users who prefer or require alternative communication methods. The system emits low-noise radio waves that pass through the face and become modulated as they cross the tongue, lips, and soft palate during silent articulation. Overcoming cross-user anatomical differences required scaling up pretraining data by collecting samples from numerous participants.
## BACKGROUND
Silent speech interfaces are devices that allow communication without vocalized sound by capturing articulatory movements using alternative sensing technologies. Subvocal recognition typically monitors physiological changes or physiological signals to reconstruct speech into digital text or audio output.