Tracking Object Positions in Reinforcement Learning: A Metric for Keypoint Detection (extended version).
Emma CramerJonas ReiherSebastian TrimpePublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- keypoint detection
- object tracking
- keypoints
- deformable objects
- object position
- partial occlusion
- multiple objects
- target object
- continuously moving
- object model
- particle filter
- real time
- partially occluded
- object motion
- articulated objects
- cluttered background
- tracking framework
- tracking objects
- d objects
- visual tracking
- function approximation
- image sequences
- object matching
- image processing
- rigid objects
- camera views
- learning algorithm
- model free
- moving objects
- mean shift
- appearance model
- feature extraction
- bounding box
- reinforcement learning algorithms
- face recognition
- robust tracking
- moving target
- object appearance
- kalman filter
- markov decision processes
- multiscale
- object location
- relative position
- optimal policy