Optical-flow features empirical mode decomposition for motion anomaly detection.
Moacir PontiTiago S. NazaréJosef KittlerPublished in: ICASSP (2017)
Keyphrases
- anomaly detection
- optical flow
- intrusion detection
- empirical mode decomposition
- image sequences
- network intrusion detection
- motion field
- feature vectors
- anomalous behavior
- moving objects
- motion estimation
- network traffic
- intrusion detection system
- motion model
- network anomaly detection
- detecting anomalies
- feature extraction
- non stationary
- negative selection algorithm
- feature space
- co occurrence
- sparse pca
- reinforcement learning
- principal component analysis
- computer vision
- normal behavior
- wavelet transform
- pattern recognition
- multiresolution