RiskWorld: Object-Centric Latent World Modeling for Autonomous Driving Risk Identification
产业动态
来源：arXiv cs.RO发布时间待核实
arXiv:2608.21414v1 Announce Type: new Abstract: Autonomous driving risk identification aims to determine which observed object is likely to become safety-critical to the ego vehicle. Existing approaches typically predict scene-level accidents, infer risk objects indirectly from ego behavior, or apply geometric checks after trajectory forecasting, without directly using predicted ego--object relations for risk-source localization.