The preliminary study of robust speech feature extraction based on maximizing the accuracy of states in deep acoustic models.
Li-chia ChangJeih-weih HungPublished in: ROCLING (2020)
Keyphrases
- preliminary study
- feature extraction
- speech recognition
- acoustic models
- broadcast news
- speech recognizer
- automatic speech recognition
- speaker independent
- face recognition
- speaker identification
- classification accuracy
- principal component analysis
- speech signal
- feature selection
- hidden markov models
- image processing
- visual features
- language model
- noisy environments
- spoken language
- video sequences
- training data
- multimedia
- computer vision