Fast Human Activity Recognition Based on a Massively Parallel Implementation of Random Forest.
Jan JanousekPetr GajdosPavel DohnálekMichal RadeckýPublished in: ACIIDS (2) (2016)
Keyphrases
- random forest
- massively parallel
- human activity recognition
- activity recognition
- random forests
- human activities
- parallel computing
- decision trees
- fine grained
- ensemble methods
- feature set
- parallel machines
- multi label
- visual surveillance
- data analysis
- text categorization
- image classification
- active learning
- reinforcement learning
- training data
- data sets
- decision tree learning algorithms