Online MVDR Beamformer Based on Complex Gaussian Mixture Model With Spatial Prior for Noise Robust ASR.
Takuya HiguchiNobutaka ItoShoko ArakiTakuya YoshiokaMarc DelcroixTomohiro NakataniPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2017)
Keyphrases
- gaussian mixture model
- noisy environments
- minimum variance
- signal to noise ratio
- mixture model
- speaker identification
- em algorithm
- feature vectors
- maximum likelihood
- noise reduction
- speaker recognition
- background subtraction
- speech recognition
- gaussian distribution
- probability density function
- gaussian mixture modeling
- expectation maximization
- gaussian mixture
- automatic speech recognition
- prior knowledge
- feature space
- mixture distribution
- computer vision
- gaussian model
- image processing