Is your feature request related to a problem? Please describe.
Currently the blink detection is completely hardcoded, it uses the distance between your top and open eye lid. Over time we have added many new blink detection modes however, we are quickly getting to a point where we can't add every mode and all the options to go along with it.
Describe the solution you'd like
Instead of hand built blink algorithms we should use some machine learning to detect when a use blinks. This would have to be trained on a per user basis but might require a base model.
Describe alternatives you've considered
If we cant build an ML model then we can go back to building it by hand
Is your feature request related to a problem? Please describe. Currently the blink detection is completely hardcoded, it uses the distance between your top and open eye lid. Over time we have added many new blink detection modes however, we are quickly getting to a point where we can't add every mode and all the options to go along with it.
Describe the solution you'd like Instead of hand built blink algorithms we should use some machine learning to detect when a use blinks. This would have to be trained on a per user basis but might require a base model.
Describe alternatives you've considered If we cant build an ML model then we can go back to building it by hand