tanning ASR is an automatic speech recognizer (ASR) integrated with several speaker adaptation algorithms. It supports CUED's HTK speech recognition toolkit acoustic model and acoustic features.
It's implemented for my work on rapid speaker adaptation. The idea is to create a set of tools which can use HTK's models and mfc files to perform my tests. (I have to say that it's quite boring to modify HTK's code ;) ). tanning-ASR mainly contains a library which implements many algorithms in the field of speech recognition, particularly for speaker adaptation. It includes also some tools to extend the functionality of HTK: hmmprint (print summary info of a hmmdefs), hmmdiff (distance measure betwen two hmmdefs) etc.