LAE-ST-MOE: Boosted Language-Aware Encoder Using Speech Translation Auxiliary Task for E2E Code-Switching ASR.
Guodong MaWenxuan WangYuke LiYuting YangBinbin DuHaoran FuPublished in: ASRU (2023)
Keyphrases
- automatic speech recognition
- speech recognition
- speech signal
- language resources
- broadcast news
- target language
- spoken language
- spontaneous speech
- text to speech
- language acquisition
- word error rate
- machine translation
- noisy environments
- machine translation system
- source language
- bit rate
- speech synthesis
- error control
- parallel corpus
- english text
- text to speech synthesis
- hidden markov models
- rate distortion
- source code
- recognition errors
- bilingual dictionaries
- natural language
- speech corpus
- language generation
- conversational speech
- comparable corpora
- language learning
- language processing
- programming language
- spoken dialog systems
- information extraction
- query translation
- error detection