What can AI/ML do with voice/audio from a video?
It may be interesting to see what AI can do about scripting instructions from JOI videos for use with haptics. Not the ones where they're jerking off a dildo, but using AI to recognize the voice commands from the girl like "faster", "just the head", "long, full strokes". Then executing those commands on the toy.
Or giving the option to remove speaking or moaning from the stunk cock.
Maybe selectively customize the pitch/timbre of the model's voice so that she's lower and more sultry or higher and chipper.
It may have been mentioned before, but what about manipulation of what the model says so that you can insert a user's name or remove any mention of "step-"? I guess if you're going that far you may as well look at changing the whole "story" of the scene by changing the model's words, but after a certain point it would look like 1970s dubbed kung fu movies.
