The mask I'm trying to build will display different lighting patterns. The overall goal is to synchronize these patterns with a musical/dance performance, so there were two options: either hard-code the sequence in the microcontroller or equip the mask with one or two sensors so the change of patterns can be triggered by the wearer.

I decided to "explore" the second option first. Tracking jaw movements (how open the mouth is) and maybe the cheeks (or lips) might make it follow the speech organically. Would a Flex sensor be a good idea for sensing (vertical) jaw movements? Note that it is supposed to be a cloth mask, so thats why I'd like to avoid making a hinge and equipping it with a potentiometer for precise angle tracking.

What could I do about tracking the cheeck/lip movent? I need only a couple of states, nothing precise. A strain sensor, perhaps?

I hope I was able to convey my question, if not please let me know

EDIT: I apologize, but I omitted the fact that it is supposed to be a portable, real-time system.

  • \$\begingroup\$ I think the "best" sensor for this would be a camera and some video processing. There are a lot of tutorials on the web about getting lip/mouth movements. Camera's are small and you can remove the IR filter and light the inside with IR LED's, focus distance might be a problem though. \$\endgroup\$ – Ron Beyer Oct 18 at 13:12
  • 1
    \$\begingroup\$ @RonBeyer I don't think that works on a cloth mask. \$\endgroup\$ – DKNguyen Oct 18 at 13:13
  • \$\begingroup\$ I thought there was fabric that changed conductivity when stretched. \$\endgroup\$ – DKNguyen Oct 18 at 13:14
  • 1
    \$\begingroup\$ What @RonBeyer is suggesting is actually not a bad idea. A lot of voice and image/video recognition today involve the Fourier Transform of a signal and then you would perform convolution on the signal to something the user defines to get a match. You can train a program learn to particular sounds and phonetics and feed it into a convolutional neural network. CMU Sphinx is a well developed open source API for voice recognition. I've used it before and it does indeed need a ton of training but there's documentations on how to train it. \$\endgroup\$ – KingDuken Oct 18 at 14:52
  • 1
    \$\begingroup\$ The new Google Pixel Buds will have an accelerometer to detect jaw movement. \$\endgroup\$ – Leon Heller Oct 18 at 16:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.