A Beamforming Algorithm Based on Maximum Likelihood of a Complex Gaussian Distribution With Time-Varying Variances for Robust Speech Recognition.
Byung Joon ChoJun-Min LeeHyung-Min ParkPublished in: IEEE Signal Process. Lett. (2019)
Keyphrases
- gaussian distribution
- speech recognition
- maximum likelihood
- expectation maximization
- noisy environments
- em algorithm
- speaker adaptation
- speech enhancement
- language model
- dimensionality reduction
- probabilistic model
- hidden markov models
- k means
- gaussian noise
- noise model
- training data
- speech synthesis
- smoothing algorithm
- computer vision
- neural network