Matbn 2002: a Mandarin Chinese Broadcast News Corpus
نویسنده
چکیده
The MATBN 2002 Mandarin Chinese broadcast news corpus contains a total of 40 hours of broadcast news from Public Television Service Foundation (Taiwan) with corresponding transcripts. The primary motivation for this collection is to provide training and testing data for continuous speech recognition evaluation in the broadcast domain. We expect to collect and process 220 hours of Mandarin Chinese broadcast news speech over 3 years. At the end of the first year, the 40 hour broadcast news corpus has been completed on schedule and is scheduled to be releasable in early 2003. According to our plan, we expect to release the interim 120 hour broadcast news corpus in late 2003 and the final 220 hour broadcast news corpus in late 2004.
منابع مشابه
MATBN: A Mandarin Chinese Broadcast News Corpus
The MATBN Mandarin Chinese broadcast news corpus contains a total of 198 hours of broadcast news from the Public Television Service Foundation (Taiwan) with corresponding transcripts. The primary purpose of this collection is to provide training and testing data for continuous speech recognition evaluation in the broadcast news domain. In this paper, we briefly introduce the speech corpus and r...
متن کاملAn Empirical Study of Word Error Minimization Approaches for Mandarin Large Vocabulary Continuous Speech Recognition
This paper presents an empirical study of word error minimization approaches for Mandarin large vocabulary continuous speech recognition (LVCSR). First, the minimum phone error (MPE) criterion, which is one of the most popular discriminative training criteria, is extensively investigated for both acoustic model training and adaptation in a Mandarin LVCSR system. Second, the word error minimizat...
متن کاملBroadcast news transcription in Mandarin
In this paper, our work in developing a Mandarin broadcast news transcription system is described. The main focus of this work is a port of the LIMSI American English broadcast news transcription system to the Chinese Mandarin language. The system consists of an audio partitioner and an HMM-based continuous speech recognizer. The acoustic models were trained on about 24 hours of data from the 1...
متن کاملOn Using Entropy Information to Improve Posterior Probability-Based Confidence Measures
In this paper, we propose a novel approach that reduces the confidence error rate of traditional posterior probability-based confidence measures in large vocabulary continuous speech recognition systems. The method enhances the discriminability of confidence measures by applying entropy information to the posterior probability-based confidence measures of word hypotheses. The experiments conduc...
متن کاملReal-time rich-content transcription of Chinese broadcast news
This paper describes the recent development of an Audio Indexing System for Chinese (Mandarin) broadcast news. Key issues of the three major components: automatic speech recognition, speaker identification and named entity extraction are addressed. The Chinese-language-specific challenges are discussed and our solutions are described. The recognition accuracy of the final system is comparable t...
متن کامل