A Speech/Music Discriminator of Radio Recordings Based on Dynamic Programming and Bayesian Networks.
Aggelos PikrakisTheodoros GiannakopoulosSergios TheodoridisPublished in: IEEE Trans. Multim. (2008)
Keyphrases
- bayesian networks
- audio recordings
- dynamic programming
- acoustic features
- music information retrieval
- audio visual
- spontaneous speech
- audio features
- speech music discrimination
- audio signals
- speech signal
- wireless communication
- automatic speech recognition
- speech recognition
- speaker verification
- inference in bayesian networks
- graphical models
- music retrieval
- structure learning
- probability distribution
- state space
- linear programming
- conditional probabilities
- greedy algorithm
- stereo matching
- visual features
- probabilistic inference
- human machine interaction
- reinforcement learning
- digital audio
- music recommendation
- speech synthesis
- gaussian mixture model
- probabilistic reasoning
- conditional independence
- single machine
- multi modal
- spoken language
- learning bayesian networks
- random variables
- text classification
- bayesian network classifiers
- noisy environments
- coarse to fine
- electromagnetic field
- dp matching
- multiply sectioned bayesian networks
- incomplete data