Is it feasible to use the speech recognition engine to record text from a movie in the form of a wave file? I just spent 3 days doing closed captions for one of my feature films. Unfortunately the actors did not follow my script very closely so I had to do a lot of adjustments to the text. I decided to try to automate that process. In perusing sample code for the speech objects, it seems that perhaps the only words recognized would have to be in the choices object. Is that a false assumption on my part?
↧
