Improving Generalization in Visual Reinforcement Learning via Conflict-aware Gradient Agreement Augmentation.
Siao LiuZhaoyu ChenYang LiuYuzheng WangDingkang YangZhile ZhaoZiqing ZhouXie YiWei LiWenqiang ZhangZhongxue GanPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- visual information
- function approximation
- visual features
- learning algorithm
- policy gradient
- visual analysis
- mobile robot
- state space
- model free
- conflict resolution
- resolving conflicts
- image gradient
- reinforcement learning algorithms
- temporal difference
- visual perception
- visual cues
- edge detection
- supervised learning
- low level
- learning process
- multi agent systems
- keywords
- computer vision