Findings could improve training for lip readers and boost the capacity for artificial intelligence to read lips and provide transcription and other digital services.

Horrifying potential for AI but I guess you do have to throw it in the mix in order to get funding for research like this

[–] artwork@lemmy.world 1 points 6 days ago

It does not seem the technologies in common use utilize anything actually serious, yet some devices are in design for the following, in also public patents:

Some disclosed embodiments involve determining an emotional state of an individual associated with the facial skin micromovements, and extracting meaning from the at least one subvocalized phoneme and the determined emotional state. The term “emotional state” refers to an individual's emotional condition and may be used as an indicator of the individual's behavior, cognition, and overall well-being...

As another example, the additional data may include the individual making a vocal statement that is different from a statement associated with their subvocal facial skin micromovements. Using such a vocal statement in the authentication is desirable to indicate that the user does not intend to make that statement at that time, such as in situations of duress like being threatened to say that statement...

The event may have an asymmetrical impact on the subject. In such cases, a sensing device may be used for monitoring/detecting muscle micromovements on both sides of the face and comparing said muscle micromovements. The comparison and differences between the facial muscle micromovements of each side of the face may be used to determine the extent of damage, as well as monitor deterioration or improvement in the subject's condition. As a difference in facial muscle micromovements may be determined and may be above a certain threshold, an indication of an illness/condition or episode may be generated.

Source: https://patents.google.com/patent/US20250173415A1/en