Real-time On-line Learning of Transformed Hidden Markov Models from Video.
Nemanja PetrovicNebojsa JojicBrendan J. FreyThomas S. HuangPublished in: AISTATS (2003)
Keyphrases
- hidden markov models
- real time
- markov models
- video sequences
- speech recognition
- visual speech recognition
- conditional random fields
- sequence classification
- audio signals
- high definition
- gesture recognition
- video data
- video content
- sequential data
- markov model
- hidden states
- continuous hidden markov models
- viterbi algorithm
- maximum entropy
- hidden state
- discriminative training
- speech signal
- dynamic scenes
- video frames
- space time
- head tracking
- stochastic context free grammars
- baum welch
- random effects
- moving objects
- image sequences