Vector-quantization based mask estimation for missing data automatic speech recognition.
Maarten Van SegbroeckHugo Van hammePublished in: INTERSPEECH (2007)
Keyphrases
- edge detection
- vector quantization
- missing data
- automatic speech recognition
- speech recognition
- image compression
- speech signal
- multiscale
- missing values
- vector quantizer
- hidden markov models
- image processing
- incomplete data
- structure from motion
- broadcast news
- distortion measure
- noisy environments
- codebook design
- finite state vector quantization
- language model
- pattern recognition
- feature extraction