Comparing Gaussian Mixture Model and Hidden Markov Model to Classify Unique Physical Activities from Accelerometer Sensor Data.
Arindam DuttaOwen MaMeynard John ToledoMatthew P. BumanDaniel W. BlissPublished in: ICMLA (2016)
Keyphrases
- hidden markov models
- sensor data
- gaussian mixture model
- human activities
- gesture recognition
- activity recognition
- sensor networks
- mixture model
- data streams
- em algorithm
- feature vectors
- sensor measurements
- maximum likelihood
- background subtraction
- speaker recognition
- expectation maximization
- speech recognition
- conditional random fields
- feature space
- wearable sensors
- health monitoring
- hidden semi markov models
- speaker identification
- gaussian distribution
- density estimation
- probability density function
- pattern recognition
- gaussian mixture
- object recognition
- information retrieval
- speech signal
- automatic speech recognition
- visual words
- active learning