Goal Oriented Opportunistic Recognition of High-Level Composed Activities Using Dynamically Configured Hidden Markov Models.
Gerold HoelzlMarc KurzAlois FerschaPublished in: ANT/MobiWIS (2012)
Keyphrases
- hidden markov models
- goal oriented
- gesture recognition
- handwritten text recognition
- human activities
- activity recognition
- phoneme recognition
- keyword spotting
- markov models
- speaker independent
- low level
- speech recognition
- conditional random fields
- visual speech recognition
- hidden state
- handwritten word recognition
- text recognition
- feature extraction
- requirements analysis
- hidden semi markov models
- pattern recognition
- sequence classification
- sequential data
- visual speech
- handwritten words
- markov model
- human motion
- automatic speech recognition systems
- handwriting recognition
- action recognition
- hierarchical hidden markov model
- continuous hidden markov models