Decent speech-to-text is a long way from being a solved problem, so automation isn't a complete solution, but it's certainly a huge assist over doing it completely manually.
A good place to start looking if you're after an OSS solution is what Carnegie Mellon University are doing - there's a variety of tools at http://cmusphinx.sourceforge.net
A good place to start looking if you're after an OSS solution is what Carnegie Mellon University are doing - there's a variety of tools at http://cmusphinx.sourceforge.net