Improved Training of End-to-end Attention Models for Speech Recognition.
Albert ZeyerKazuki IrieRalf SchlüterHermann NeyPublished in: INTERSPEECH (2018)
Keyphrases
- end to end
- speech recognition
- acoustic models
- hidden markov models
- pattern recognition
- speech synthesis
- wall street journal corpus
- isolated word
- speech signal
- speaker independent
- language model
- speech recognition technology
- noisy environments
- speech processing
- speech recognizer
- computer vision
- speaker identification
- speaker diarization
- speech recognition systems
- computational complexity
- audio visual speech recognition