Hey all,I'm building a web platform using express.js which needs to render words per minute in time in form of a graph. The input stream would be a user's mic device.I'm currently reading on HMM and STFT in order to be able to extract the words from the stream. But the implementation seems to require a bit more audio processing knowledge that i can grasp at this point. Of course, i need here a solution which doesn't need to be very accurate, but just to approximate the values in time with some [minimal if possible] margin of error.Are there any good node libraries that i can use in order to ease the implementation ? Any ready to use package [binary] that can provide such info ? not necessary need to be nodejs. Since i came across the Praat, but it's scripting language seems like another layer of learning on top of dsp.Any pointers on how this can be done without too much hassle would be more than welcome.Thanx in advance ..
Submitted June 02, 2017 at 02:19PM by mishkez
No comments:
Post a Comment