Building A Highly Accurate Mandarin Speech Recognizer With Language-Independent Technologies and Language-Dependent Modules.
Mei-Yuh HwangGang PengMari OstendorfWen WangArlo FariaAaron HeidelPublished in: IEEE Trans. Speech Audio Process. (2009)
Keyphrases
- highly accurate
- language independent
- speech recognizer
- speech recognition
- broadcast news
- spoken language
- chinese text retrieval
- parallel corpus
- n gram
- machine translation
- language processing
- language model
- word meanings
- high accuracy
- text retrieval
- cross lingual
- text classification
- word level
- high quality
- cross language
- target language
- natural language
- automatic speech recognition
- speech signal
- data mining
- hidden markov models
- metadata
- dialogue system
- word segmentation
- source language
- text mining
- parallel corpora
- machine learning