Training Wideband Acoustic Models Using Mixed-Bandwidth Training Data for Speech Recognition.
Michael L. SeltzerAlex AceroPublished in: IEEE Trans. Speech Audio Process. (2007)
Keyphrases
- acoustic models
- speech recognition
- speech signal
- speech recognizer
- hidden markov models
- automatic speech recognition
- speaker independent
- language model
- pattern recognition
- training data
- noisy environments
- broadcast news
- discriminative training
- training process
- speech recognition systems
- n gram
- multimedia
- machine learning
- speaker identification
- acoustic features
- character recognition