CONF Dines_INTERSPEECH_2009/IDIAP Measuring the gap between HMM-based ASR and TTS Dines, John Yamagishi, Junichi King, Simon speech recognition speech synthesis unified models EXTERNAL https://publications.idiap.ch/attachments/papers/2009/Dines_INTERSPEECH_2009.pdf PUBLIC https://publications.idiap.ch/index.php/publications/showcite/Dines_Idiap-RR-16-2009 Related documents Proceedings of Interspeech Brighton, U.K. 2009 September 2009 The EMIME European project is conducting research in the development of technologies for mobile, personalised speech-tospeech translation systems. The hidden Markov model is being used as the underlying technology in both automatic speech recognition (ASR) and text-to-speech synthesis (TTS) components, thus, the investigation of unified statistical modelling approaches has become an implicit goal of our research. As one of the first steps towards this goal, we have been investigating commonalities and differences between HMM-based ASR and TTS. In this paper we present results and analysis of a series of experiments that have been conducted on English ASR and TTS systems measuring their performance with respect to phone set and lexicon, acoustic feature type and dimensionality andHMM topology. Our results show that, although the fundamental statistical model may be essentially the same, optimal ASR and TTS performance often demands diametrically opposed system designs. This represents a major challenge to be addressed in the investigation of such unified modelling approaches. REPORT Dines_Idiap-RR-16-2009/IDIAP Measuring the gap between HMM-based ASR and TTS Dines, John Yamagishi, Junichi King, Simon EXTERNAL https://publications.idiap.ch/attachments/reports/2009/Dines_Idiap-RR-16-2009.pdf PUBLIC Idiap-RR-16-2009 2009 Idiap July 2009 The EMIME European project is conducting research in the development of technologies for mobile, personalised speech-tospeech translation systems. The hidden Markov model is being used as the underlying technology in both automatic speech recognition (ASR) and text-to-speech synthesis (TTS) components, thus, the investigation of unified statistical modelling approaches has become an implicit goal of our research. As one of the first steps towards this goal, we have been investigating commonalities and differences between HMM-based ASR and TTS. In this paper we present results and analysis of a series of experiments that have been conducted on English ASR and TTS systems measuring their performance with respect to phone set and lexicon, acoustic feature type and dimensionality andHMM topology. Our results show that, although the fundamental statistical model may be essentially the same, optimal ASR and TTS performance often demands diametrically opposed system designs. This represents a major challenge to be addressed in the investigation of such unified modelling approaches.