Human detection in MOUT scenarios using covariance descriptors and supervised manifold learning.
Jürgen MetzlerDieter WillersinnPublished in: Visual Information Processing (2010)
Keyphrases
- manifold learning
- human detection
- semi supervised
- action recognition
- surveillance system
- object detection
- low dimensional
- dimensionality reduction
- high dimensional
- pedestrian detection
- dimension reduction
- feature extraction
- video surveillance
- unsupervised learning
- supervised learning
- background subtraction
- pairwise
- semi supervised learning
- learning algorithm
- high dimensional data
- feature selection
- computer vision
- feature space
- machine learning
- human body
- image processing
- sparse representation
- principal component analysis
- pattern recognition
- nearest neighbor
- partial occlusion
- data points