Applying noise compensation methods to robustly predict acoustic speech features from MFCC vectors in noise.
Ben MilnerJonathan DarchSaeed VaseghiPublished in: ICASSP (2008)
Keyphrases
- noisy data
- feature extraction
- speech recognition systems
- noise sensitivity
- speech recognition
- feature vectors
- classification accuracy
- discriminatory power
- feature set
- signal to noise ratio
- noise level
- image statistics
- noisy environments
- noise reduction
- missing data
- benchmark datasets
- image features
- feature descriptors
- median filter
- hybrid method
- gaussian mixture model
- classification method
- facial expressions