Adaptive visual tracking using the prioritized Q-learning algorithm: MDP-based parameter learning approach
作者：
Highlights：
• We use an MDP formulation for optimal adaptation of tracking algorithms.
• We optimize the tracker control parameters using prioritized Q-learning.
• The proposed prioritized Q-learning approach is based on sensitivity analysis.
• The performance of our method is superior to other approaches.
• The proposed method can balance tracking accuracy and speed.
摘要
•We use an MDP formulation for optimal adaptation of tracking algorithms.•We optimize the tracker control parameters using prioritized Q-learning.•The proposed prioritized Q-learning approach is based on sensitivity analysis.•The performance of our method is superior to other approaches.•The proposed method can balance tracking accuracy and speed.
论文关键词：Adaptive visual tracking,Prioritized Q-learning,Markov decision process,Dynamic parameter optimization
论文评审过程：Received 12 August 2013, Revised 7 March 2014, Accepted 28 August 2014, Available online 4 September 2014.
论文官网地址：https://doi.org/10.1016/j.imavis.2014.08.009
