A Survey on Knowledge Graph-Based Methods for Automated Driving

Deep learning methods have made remarkable breakthroughs in machine general and automated driving (AD) particular. However, there are still unsolved problems to guarantee reliability safety of systems, especially effectively incorporate all available information knowledge the task. Knowledge graphs (KG) recently gained significant attention from both industry academia for applications that benefit by exploiting structured, dynamic, relational data. The complexity graph-structured data with complex relationships inter-dependencies between objects has posed challenges existing algorithms. recent progress graph embeddings neural networks allows applying Therefore, we motivate discuss KGs applied AD. Then, survey, analyze categorize ontologies KG-based approaches We current research propose promising future directions solutions