Tracking and classification of arbitrary objects with bottom-up/top-down detection.
Michael HimmelsbachHans-Joachim WünschePublished in: Intelligent Vehicles Symposium (2012)
Keyphrases
- crowded scenes
- partial occlusion
- robust detection
- bounding box
- classification accuracy
- human detection
- pattern recognition
- multiple objects
- machine learning
- multi target tracking
- complex background
- support vector machine svm
- d objects
- real time
- cluttered background
- support vector
- object detection
- detecting objects
- decision trees
- deformable objects
- detection rate
- false positives
- support vector machine
- tracking of multiple objects
- object detection and tracking
- object segmentation
- feature vectors
- image classification
- feature extraction
- man made
- particle filter
- object location
- feature selection
- detecting and tracking multiple
- multi object tracking
- supervised learning
- tracking objects
- text classification
- cluttered scenes
- partially occluded
- detection algorithm
- moving target
- feature space
- particle filtering
- probabilistic model
- detection method