Chapter 1 Quantifying the Importance of Recognizing Tones

  • Published 2007

Abstract

Some of the results reported here are akin to those in earlier work (Surendran & Levow (2004a), Surendran & Niyogi (2003), Surendran (2003)) but they were done here with improved computational techniques and a larger dataset. All calculations in this chapter were done using an automatically transcribed corpus of 949 111 words in 151 940 sentences from the Mandarin VOA TDT 2 collection. It contained 1020 syllable tokens and 19 788 word tokens.

48 Figures and Tables

Cite this paper

@inproceedings{2007Chapter1Q, title={Chapter 1 Quantifying the Importance of Recognizing Tones}, author={}, year={2007} }