This content has been marked as final. Show 3 replies
did you figure it out? i'm trying to write the front end to the program with VQ as my method but i'm having a little trouble.
not really ...you need at first to apply a noise reduction algorithm to audio stream[low pass filter]then apply it to neural network algorithm [back propagation]for training and testing...but the thing i cannot figure out till now is what would neuron object will look like...is it a single bit or part of the stream or what?...but anyway it can be easily done using matlab
Have you seen jsapi documentation. java doesnt used to identify the speech of particular user.
it only used for identifying spoken text.