Comparison of DCT and autoencoder-based features for DNN-HMM multimodal silent speech recognition.
Licheng LiuYan JiHongcui WangBruce DenbyPublished in: ISCSLP (2016)
Keyphrases
- speech recognition
- hidden markov models
- speech recognition systems
- cepstral coefficients
- speech synthesis
- automatic speech recognition
- speech signal
- language model
- speech recognition technology
- pattern recognition
- speaker independent
- speech processing
- speech recognizer
- noisy environments
- feature set
- handwriting recognition
- speaker dependent
- classification accuracy
- low level
- feature space
- multi modal
- speaker recognition
- mel frequency cepstral coefficients
- keyword spotting
- speaker adaptation
- speech retrieval
- feature extraction
- speech recognizers
- speaker identification
- extracting features
- image compression
- feature vectors