Tracking object positions in reinforcement learning: A metric for keypoint detection.
Emma CramerJonas ReiherSebastian TrimpePublished in: L4DC (2024)
Keyphrases
- reinforcement learning
- keypoint detection
- keypoints
- deformable objects
- object tracking
- object position
- partial occlusion
- continuously moving
- multiple objects
- d objects
- cluttered background
- state space
- target object
- function approximation
- metric space
- articulated objects
- particle filter
- tracking objects
- temporal difference
- tracking framework
- kalman filter
- partially occluded
- rigid objects
- reinforcement learning algorithms
- learning algorithm
- object classes
- machine learning
- multiscale
- position information
- moving objects
- real time
- motion analysis
- visual tracking
- object model
- markov decision processes
- appearance model
- image features
- image processing
- object matching
- multi agent
- motion model
- object appearance
- lighting conditions
- face recognition
- particle filtering
- position and orientation
- bounding box