End-to-end Tibetan Ando dialect speech recognition based on hybrid CTC/attention architecture.
Jingwen SunGang ZhouHongwu YangMan WangPublished in: APSIPA (2019)
Keyphrases
- end to end
- speech recognition
- hidden markov models
- language model
- admission control
- pattern recognition
- automatic speech recognition
- speech recognizer
- speech signal
- noisy environments
- speech recognition systems
- congestion control
- speech processing
- speech recognition technology
- speech synthesis
- speaker identification
- application layer
- speaker dependent
- speech recognizers
- audio visual speech recognition
- multi stream
- speaker independent
- speech retrieval
- image coding
- isolated word
- text localization and recognition