I am making a speech based web browser. I need it to recognize words from a list of about 50 isolated English words spoken in Indian accent. Should I use the acoustic model or the language model to train it? The words would be like, "Sports","Headlines","Back".
Should I use the acoustic model or the language model to train it? The words would be like, "Sports","Headlines","Back".
You need to make both acoustic model and a language model (or grammar).