We are looking for an api to get voice to text. In our case, we want to add audiomining to video files, which means we want to automatically generate tagwords to the video and give the user the chance to jump directly to the timecode where the tagwords are spoken.
I found the Google Speech API which seems to work quit good, but the documentation under http://lists.w3.org/Archives/Public/public-xg-htmlspeech/2011Feb/att-0020/api-draft.html is not the best, and we didn t found a way yet to trigger the start and stop record event automatically (it ends after system thinks the input is over). Even it sounds like the system is not ready for that case…
I also found this post https://stackoverflow.com/questions/2080401/is-there-a-speech-to-text-api-by-google here, but it seems like it is only possible on android systems.
So basically my question is: Is there a away to use the Google Speech API with something like flash or PHP/JS (and if yes are there any good examples) and if not does anyone know some other API with some good documentation or example codes to get voice in video to text?
Thanks,
kris
Answer to myself: Seems like, there s no way to work with the Google Speech API on Web Application as a free speech recognition engine yet. At the moment Google uses it for their own use. Hope they ll change it soon 😉
We are using Microsoft Speech API (SAPI) yet. Not the best results but ok.