It is very clear that new applications appear every day for glasses, since they are easy to use and sensors can be placed on them that significantly expand their functions. They no longer only serve to see, every day new possible functions are added to them.
Google was the first company to try something serious and different with its Google Glasses almost 10 years ago, but others have continued to try to find new applications for them, apart from augmented reality.
I have recently seen some interesting information about a development carried out at Cornell University, and it is about glasses capable of detecting oral instructions based on the movement of the lips and mouth.
The glasses (called EchoSpeech) have two microphones and two speakers that work like a portable sonar, sending sound waves to the face and being able to detect the movements of the mouth. And then with the help of an AI program it transforms those gestures into the corresponding "spoken" instructions.
In the tests carried out by the researchers, and with a deep learning algorithm, they have achieved an accuracy of 95% of the instructions transmitted in a few minutes.
The acoustic detection system consumes very little (much less than the video from a camera) with which the battery lasts much longer, it is small (then it does not weigh), and it can be easily placed on any glasses.
To work, the glasses are connected to a mobile via Bluetooth where the AI program is and in this way all the information generated is kept under control in real time and without the need to access any type of computer cloud, and this also maintains the total privacy of what is done.
The practical applications can be varied. From giving instructions in environments that require silence, giving voice to people who have lost it, or others that may appear.
I think it's a good idea and we just have to hope that the developers are able to put a product on the market that uses this technology.
EN 
ES