A Time-Synchronous Phonetic Decoder For A Long-Contextual-Span Hidden Trajectory Model

A novel time-synchronous decoder, designed specifically for a Hidden Trajectory Model (HTM) whose likelihood score computation depends on long-span phonetic contexts, is presented. HTM is a recently developed acoustic model aimed to capture the underlying dynamic structure of speech coarticulation and reduction using a compact set of parameters. The long-span nature of the HTM had posed a great technical challenge for developing efficient search algorithms for full evaluation of the model. Taking on the challenge, the decoding algorithm is developed to deal effectively with the exponentially increased search space by HTM-specific techniques for hypothesis representation, word-ending recombination, and hypothesis pruning. Experimental results obtained on the TIMIT phonetic recognition task are reported, extending our earlier HTM evaluation paradigms based on N-best and A* lattice rescoring.