Mandarin tone recognition using affine-invariant prosodic features and tone posteriorgram
نویسندگان
چکیده
This paper proposes to analyze the normalization schemes for prosodic features in terms of the affine invariance property, and shows that better robustness of the prosodic features across different prosodic conditions can be obtained in this way. The analysis is in good consistency with experimental results for Mandarin tone recognition, in which the use of both syllable-level mean and utterance-level standard deviation for pitch feature normalization offers the best recognition accuracy. Also, we incorporate tone posteriorgrams in the second-pass tone recognition, and further improved tone recognition accuracy was obtained.
منابع مشابه
Improved large vocabulary Mandarin speech recognition by selectively using tone information with a two-stage prosodic model
The incorporation of prosodic information in large vocabulary continuous speech recognition has attracted much attention in recent years, especially for a tonal language such as Mandarin Chinese. The tones of some syllables are very difficult to recognize correctly due to the very complicated prosodic behavior. Tone recognition errors inevitably degrade the recognition accuracy seriously. We pr...
متن کاملProsodic modeling for improved speech recognition and understanding
The general goal of this thesis is to model the prosodic aspects of speech to improve humancomputer dialogue systems. Towards this goal, we investigate a variety of ways of utilizing prosodic information to enhance speech recognition and understanding performance, and address some issues and difficulties in modeling speech prosody during this process. We explore prosodic modeling in two languag...
متن کاملAssessing context and learning for isizulu tone recognition
Prosody plays an integral role in spoken language understanding. In isiZulu, a Nguni family language with lexical tone, prosodic information determines word meaning. We assess the impact of models of tone and coarticulation for tone recognition. We demonstrate the importance of modeling prosodic context to improve tone recognition. We employ this less commonly studied language to assess models ...
متن کاملProsodic structure in language understanding: evidence from tone sandhi in Mandarin.
Two experiments show that prosodic information plays a crucial role in the processing of sentences of Standard Mandarin Chinese, where local lexical ambiguities may occur due to the operation of a tone sandhi rule. In Chinese, each word is associated with a tone; in this paper, the term "Mandarin tone sandhi" refers to a phonological rule that changes the first of two consecutive low tones (Ton...
متن کاملA preliminary study on acoustic correlates of tone2+tone2 disyllabic word stress in Mandarin
This paper investigated the potential acoustic correlates of word stress within a disyllabic tonal sequence, a rising tone followed by a rising tone (Tone2 Tone2) in Mandarin, based on a large corpus with adequate information of stress patterns and prosodic boundary levels. The results showed that a) For Tone2+Tone2 words, features based on tone nucleus were more effective than that of the whol...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2010