发表机构
AFARI World Model Team（AFARI世界模型团队）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
BehaviorWorldGen通过BehaviorFlow实现可控行为感知结构化世界生成，闭合驾驶动作模型与世界模拟器的循环，在多项任务上取得改进，尤其对困难交互场景效果显著。
AI 中文摘要
现代驾驶动作模型在自我改进循环中不断提升，其中学习得到的世界模拟器会想象未来观测结果，生成的数据会反馈用于优化动作模型。然而该循环的瓶颈在于模拟器无法生成周围智能体的行为合理响应，导致生成数据在交互中不真实且分布不平衡。我们提出BehaviorWorldGen，这一框架通过可控行为感知结构化世界生成闭合动作模型与世界模拟器之间的循环。其核心组件是BehaviorFlow，一种元动作条件下的交通流模型，可注入可解释的行为控制并联合生成多智能体轨迹。BehaviorFlow可实现指定的智能体行为，同时允许周围车辆对自车及彼此做出响应。生成的轨迹由世界模拟器渲染为真实的多视角观测结果，这些结果与修正后的交互感知轨迹配对以用于动作模型优化。由于BehaviorWorldGen使用结构化轨迹作为模块间的接口，它兼容多种动作模型和世界模拟器。在世界生成、场景外推和策略优化上的实验显示出持续的改进，最大收益集中在困难的交互场景中。
英文摘要
Modern driving action models are increasingly improved in a self-improvement loop, where a learned world simulator imagines future observations and the resulting data is fed back to refine the action model. However, the bottleneck of this loop lies in the simulators' inability to generate behaviorally plausible responses by surrounding agents, making generated data both unrealistic in interaction and imbalanced in distribution. We introduce BehaviorWorldGen, a framework that closes the loop between action models and world simulators through controllable behavior-aware structured world generation. Its core component is BehaviorFlow, a meta-action-conditioned traffic-flow model that injects interpretable behavior controls and jointly generates multi-agent rollouts. BehaviorFlow realizes the specified agent behaviors while allowing surrounding vehicles to respond to the ego and to one another. The resulting rollouts are rendered by a world simulator into realistic multi-view observations, which are paired with corrected interaction-aware trajectories for action-model refinement. Since BehaviorWorldGen uses structured trajectories as the interface between its modules, it is compatible with diverse action models and world simulators. Experiments on world generation, scene extrapolation, and policy refinement demonstrate consistent improvements, with the largest benefits concentrated on difficult interactive scenarios.
