Additive and convolutional noise canceling in speaker verification using a stochastic weighted viterbi algorithm.
Néstor Becerra YomaMiguel Villar FernandezPublished in: INTERSPEECH (2001)
Keyphrases
- speaker verification
- viterbi algorithm
- noisy environments
- hidden markov models
- dynamic programming
- noise reduction
- speech recognition
- markov model
- n gram
- audio visual
- speech signal
- emotion recognition
- pattern recognition
- signal to noise ratio
- automatic speech recognition
- multilayer perceptron
- neural network
- image classification
- video sequences
- reinforcement learning
- image segmentation
- information retrieval