论文

SceneAgent:跨游戏引擎可迁移的多智能体场景生成框架

SceneAgent: A Multi-Agent Framework for Portable Game Scene Generation

精选理由

做游戏或 3D 场景生成的开发者可以看看这篇:SceneAgent 把场景状态、构建经验做成可跨引擎迁移的技能库,换个引擎不用重写整套流程。有 GameSceneBench 的 69 个提示词可复现,和单智能体方案比在场景完整度上提升明显。

北京航空航天大学等机构的研究团队(arXiv 2611.17821)发布 SceneAgent,一个面向游戏场景生成的多智能体框架,核心目标是让场景构建流程和经验在不同游戏引擎之间复用。团队同时推出 GameSceneBench,包含 69 个室内外场景提示词,SceneAgent 在统一资产与评测协议下,对五个基线取得 78.2%-89.4% 的人类平局调整偏好分。

正文

架构上,框架用持久化的 SceneContext 保存演化中的场景世界状态,通过后端适配器隔离坐标系约定、资产绑定和导出逻辑;结构化交接让新的 Builder 会话能查询持久状态而不必继承完整构建记录。跨代际层面,经人工批准的构建轨迹会被蒸馏成不含资产 ID 的 Layout 技能与语义化 Content 技能,供检索和重组复用。

消融实验显示,可复用技能与多智能体协议主要提升了提示词对齐度和场景完整度,代价是额外运行时间。在第二个引擎中部署时,共享的构建协议与技能库得以保留,只需替换后端集成组件,验证了将持久状态、任务推理与可复用经验分层的设计。

原文 · arXiv cs.AI

SceneAgent: A Multi-Agent Framework for Portable Game Scene Generation

Moving a scene generator to a new game engine requires more than exporting its final output: planning, construction tools, and stored experience may also depend on the original backend. We present SceneAgent, a multi-agent framework for reusing the construction process and its experience across backends. A canonical SceneContext maintains the evolving world, while backend adapters isolate coordinate conventions, asset bindings, and export. Structured handoffs let fresh Builder sessions query this persistent state without inheriting the full construction transcript. Across generations, human-approved traces are distilled into asset-ID-free Layout skills and semantic Content skills that can be retrieved and recombined. We introduce GameSceneBench, comprising 69 indoor and outdoor prompts. Under a common asset and evaluation protocol, SceneAgent obtains human tie-adjusted preference scores of 78.2-89.4% against five baselines. Ablations show that reusable skills and the multi-agent protocol primarily improve prompt alignment and scene completeness, with additional runtime. Deployment in a second engine preserves the shared construction protocol and skill library while replacing backend integration components. Together, these results support separating persistent scene state, task-local reasoning, and reusable experience for portable scene construction.

产业坐标与信源核验

本条情报归属 AITOP「论文」核心栏目,重点聚焦中国 AI 产业与国产大模型生态。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:SceneAgent、GameSceneBench、多智能体、游戏开发,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。

信源核验说明:本文原始线索采编自「arXiv cs.AI」(@Xingyi Li, Jiahao Yang, Yiqin Zhu, Jingwen Ye, Yang Sun, Weidong Zhang),经 AITOP 多维评分管线结构化提炼。