Ultrasound Enables Silent Speech, Model Trains in a Month

akilian · x · 2026-07-08

The @alephneuro team demonstrated an ultrasound technology for "silent speech," allowing a model to recognize a user's speech intent without them making a sound. Reportedly, the model achieved impressive results after training on just one month of collected data. This research falls under silent voice decoding and neural interfaces.

Related event: Ultrasound Tech Enables Silent Speech Decoding(2 posts)→

Original post →

More from Research

Research channel →