Mondrian hidden Markov model for music signal processing.
Masahiro NakanoYasunori OhishiHirokazu KameokaRyo MukaiKunio KashinoPublished in: ICASSP (2014)
Keyphrases
- hidden markov models
- signal processing
- audio signals
- image processing
- music information retrieval
- sequential data
- speech recognition
- viterbi algorithm
- pattern recognition
- markov models
- computer vision
- conditional random fields
- hidden state
- discriminative training
- hidden states
- handwritten word recognition
- markov model
- gesture recognition
- automatic speech recognition
- hierarchical hidden markov model
- forward backward
- continuous hidden markov models
- maximum likelihood
- phoneme recognition
- baum welch
- multi stream
- feature extraction