Am currently working on a tone based language in which tone information is critical to resolve speech classification problem. The language has a CV and V syllabic structure. Hence have decided to adopt both tone and word n-gram and syllable and tone n-gram in developing the ASR system. To accomplish this with HTK, I don't know what to do. But I know how to compute the n-gram for word, tone and syllable in MATLAB environment. How do I make use of USER computed n-gram in HTK. Does anyone have answer to this problem?

--- (Edited on 3/16/2014 7:28 am [GMT-0500] by asosimi) ---