What it is: Subvocal Recognition refers to technology that transcribes human speech using silent (subvocal) muscular movements that accompany the voice but don’t produce sound in and of themselves. This creates the opportunity for interfaces that allow silent, hands-free communication using the vocal apparatus.
What it does: Speech is the result of complex muscle movements, not all of which generate sound. Thus, using myography (the measurement of muscle contraction) it’s possible to transcribe the intended communication of a “speaker” who is silently enunciating words using the mouth and throat. The process is analogous to lip-reading, but with a greater degree of sophistication, and a greater degree of privacy.
Why it matters: Our environments frequently get in the way of vocal communication. For example, it would be handy to be able to take calls on a crowded bus or in an airport, but noise and social considerations are prohibitive. Also, many would enjoy the ability to narrate their thoughts and make vocal notes they move about their day, but here, too, there are practical and social difficulties. With Subvocal Recognition, such things become possible, allowing some of the advantages of vocal communication to be married to the privacy and portability of digital communication. Additionally, future devices may allow the vocally disabled to communicate more fluidly.
What to do about it: Subvocal Recognition is a technology in its infancy. However, it’s worth keeping an eye on; when the technology matures, it will yield devices that could significantly disrupt fields such as teleconferencing, personal communication, and even smart buildings.
- Allows for “vocal” communication in noisy spaces
- Devices could have little to no learning curve given that they take advantage of existing muscle movements
- Based on well-known linguistic research about subvocal components of speech
- As of yet, prototypes are bulky and perform inadequately
- Could lead to privacy concerns in the future if devices learn to recognize subvocal communications from a distance
- Devices may have trouble adapting to distinctive movements of different speakers
- Silent vocal messaging for crowded spaces
- Silent hands-free writing/note-taking
- Speech devices for sufferers of ALS and other vocally disabling conditions
- IoT devices allowing users to silently “vocalize” with elements of the environment
Early prototypes of Subvocal Recognition technology have been around since 2008, when Texas Instruments demonstrated a somewhat low-powered device. More recently, developers at MIT created AlterEgo, a more mature subvocal headset/myograph that’s much closer to a potential consumer application. As of yet, major players in the vocal communication industry haven’t publicly joined in, but it’s possible that other prototypes are in development.