Multilingual Sequence-to-Sequence Speech Recognition: Architecture, Transfer Learning, and Language Modeling.
Jaejin ChoMurali Karthick BaskarRuizhi LiMatthew WiesnerSri Harish MallidiNelson YaltaMartin KarafiátShinji WatanabeTakaaki HoriPublished in: SLT (2018)
Keyphrases
- speech recognition
- language model
- language modeling
- transfer learning
- cross lingual
- hidden markov models
- pattern recognition
- automatic speech recognition
- probabilistic model
- information retrieval
- machine learning
- query expansion
- document retrieval
- word error rate
- word segmentation
- retrieval model
- n gram
- mixture model
- image processing
- active learning
- speech signal
- prior knowledge
- reinforcement learning