Hidden Markov Model Variational Autoencoder for Acoustic Unit Discovery.
Janek EbbersJahn HeymannLukas DrudeThomas GlarnerReinhold Haeb-UmbachBhiksha RajPublished in: INTERSPEECH (2017)
Keyphrases
- hidden markov models
- speaker independent
- speech recognition
- sequential data
- gesture recognition
- visual speech
- noisy speech
- viterbi algorithm
- markov model
- markov models
- speech recognition systems
- handwritten word recognition
- hidden states
- optical flow
- continuous hidden markov models
- dynamic probabilistic networks
- discriminative training
- speech signal
- conditional random fields
- markov random field
- image segmentation
- hidden state
- free energy
- baum welch
- automatic speech recognition
- pattern discovery