Development of an HMM-based speech synthesis system for Indian English language

This paper describes speech synthesis using HMM-based speech synthesis system (HTS) for the Indian English language. In this synthesis method, trajectories of speech parameters are generated from the trained hidden Markov models (HMM). A final speech waveform is synthesized from those speech parameters. In our experiment, context dependent HMM´s were trained over a set of 1000 phonetically rich studio recorded sentences in Indian English language. The contextual information included the positional information of a phoneme with respect to its position in word, syllable and the sentence. Experimental evaluation depicts that the developed text-to-speech system is capable of producing adequately natural speech in terms of intelligibility and intonation. Very encouraging results were obtained even with a small database size.