SpeechBERT: An Audio-and-Text Jointly Learned Language Model for End-to-End Spoken Question Answering.
Yung-Sung ChuangChi-Liang LiuHung-yi LeeLin-Shan LeePublished in: INTERSPEECH (2020)
Keyphrases
- end to end
- language model
- question answering
- information retrieval
- speech recognition
- passage retrieval
- language modeling
- text summarization
- document retrieval
- n gram
- speech transcripts
- text retrieval
- retrieval model
- text mining
- query expansion
- broadcast news
- test collection
- information extraction
- probabilistic model
- multimedia
- cross language
- query terms
- sentence retrieval
- visual information
- named entities
- natural language processing
- relation extraction
- natural language
- speech signal
- vector space model
- information retrieval systems
- automatic speech recognition
- web documents
- translation model
- relevant documents
- pseudo relevance feedback
- retrieval systems
- search engine