An SVM approach for activity recognition based on chord-length-function shape features.
Samy SadekAyoub Al-HamadiBernd MichaelisUsama SayedPublished in: ICIP (2012)
Keyphrases
- activity recognition
- shape features
- feature vectors
- human activity recognition
- human activities
- shape descriptors
- action recognition
- shape matching
- support vector machine
- accelerometer data
- smart home
- texture features
- activities of daily living
- support vector
- feature selection
- activity analysis
- machine learning
- knn
- feature space
- training data
- color features
- invariant moments
- transfer learning
- training set