发表机构
CUHK-Shenzhen; SLAI; Tsinghua SIGS; Voyager Research, Didi Chuxing; USTC（香港中文大学（深圳）; 深圳人工智能研究院; 清华大学深圳国际研究生院; 滴滴出行 Voyager 研究院; 中国科学技术大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出分布式多人世界模型WorldCast，通过本地客户端与相机对齐的玩家状态场、共享场景状态实现高效多人生成，在《反恐精英2》上验证了其一致性、实时性与可扩展性。
AI 中文摘要
多人世界模型必须生成具有一致表示的独立可控视图，涵盖玩家及其共享环境。现有多数方法通过联合多视图生成协调多个玩家，其计算成本随玩家数量增加而增长。本文提出WorldCast，一种分布式多人世界模型，每个玩家运行包含视频生成器和状态模型的本地客户端。训练时利用记录的玩家位置和地图几何结构，状态模型从生成的视频和控制输入估计玩家位置。客户端交换玩家状态并将其投影到与相机对齐的玩家状态场中，以此指导其他玩家的渲染位置与方式。共享场景状态使客户端可复用彼此生成的观测结果，从而在各视图间维持一致的场景外观。在《反恐精英2》（Counter-Strike 2）上的实验表明，WorldCast具备一致性、实时性能与分布式可扩展性：与联合生成方法相比，相机对齐的玩家状态场将玩家渲染速率提升了一个数量级以上；共享场景状态则提升了整局游戏的视觉一致性。每个客户端可实时运行，仅交换玩家与场景状态，无需集中式计算瓶颈即可实现可扩展的多人生成，且图像质量在长达数小时的推演中保持稳定。
英文摘要
Multiplayer world models must generate independently controlled views with consistent representations of both players and their shared environment. Most existing approaches coordinate multiple players through joint multi-view generation, whose cost grows with each additional player. We present WorldCast, a distributed multiplayer world model in which each player runs a local client comprising a video generator and a state model. Using recorded player positions and map geometry during training, the state model estimates the player's position from generated video and control inputs. Clients exchange player states and project them into camera-aligned player state fields that guide where and how other players are rendered. Shared scene state enables clients to reuse one another's generated observations to maintain consistent scene appearance across views. Experiments on Counter-Strike 2 demonstrate WorldCast's consistency, real-time performance, and distributed scalability. The camera-aligned player state field improves player rendering rates by over an order of magnitude over joint-generation methods, while shared scene state improves visual consistency over whole rounds. Each client runs in real time and exchanges only player and scene states, enabling scalable multiplayer generation without a centralized computational bottleneck. Image quality remains stable over hour-long rollouts.
Comments31 pages, 19 figures, 20 tables. Project page: https://ziyang-ye.github.io/WorldCast-Page
