Self-Supervise Reinforcement Learning Method for Vacant Parking Space Detection Based on Task Consistency and Corrupted Rewards.
Manh-Hung NguyenTzu-Yin ChaoChing-Chun HsiaoYung-Hui LiChing-Chun HuangPublished in: IEEE Trans. Intell. Transp. Syst. (2024)
Keyphrases
- reinforcement learning
- detection method
- detection algorithm
- high accuracy
- preprocessing
- clustering method
- significant improvement
- model free
- dynamic programming
- false positive rate
- high precision
- support vector machine svm
- temporal difference
- function approximation
- segmentation method
- clustering algorithm
- probabilistic model
- cost function
- similarity measure