Have best of both worlds: two-pass hybrid and E2E cascading framework for speech recognition.
Guoli YeVadim MazalovJinyu LiYifan GongPublished in: CoRR (2021)
Keyphrases
- speech recognition
- hidden markov models
- automatic speech recognition
- pattern recognition
- speech processing
- speech synthesis
- signal processing
- speech signal
- noisy environments
- speech recognizer
- speech understanding
- probabilistic model
- em algorithm
- noise reduction
- speaker identification
- audio visual speech recognition