arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.08751cs.AI
收藏

通过大型语言模型在 Minecraft 中生成 3D 建筑

3D Building Generation in Minecraft via Large Language Models

Southern University of Science and Technology(南方科技大学)

Shiying Hu, Zengrong Huang, Chengpeng Hu, Jialin Liu

AI总结:

本文提出 T2BM 模型,通过提示词优化、层间表示解码与修复,让 LLM 在 Minecraft 中生成结构完整、含立面与室内设施的三维建筑,实验验证了其有效性与应用潜力。

摘要:

近来,程序化内容生成借助大型语言模型(LLM),在《Super Mario Bros.》和《Sokoban》等二维游戏关卡生成领域取得了显著进展。为进一步验证 LLM 的能力,本文探究 LLM 如何助力沙盒游戏 Minecraft 中的三维建筑生成。我们提出 Text to Building in Minecraft(T2BM)模型,该模型包含提示词优化、层间表示解码与修复三个环节,支持生成建筑立面、室内场景以及门等功能方块。实验评估了通过 LLM 生成建筑的完整性与满意度。结果表明,LLM 在三维建筑生成方面具有巨大潜力:在合适的提示词下,LLM 能够在 Minecraft 中生成结构完整的正确建筑,并加入窗户、床等特定建筑方块,满足人类用户指定的需求。

英文摘要:

Recently, procedural content generation has exhibited considerable advancements in the domain of 2D game level generation such as Super Mario Bros. and Sokoban through large language models (LLMs). To further validate the capabilities of LLMs, this paper explores how LLMs contribute to the generation of 3D buildings in a sandbox game, Minecraft. We propose a Text to Building in Minecraft (T2BM) model, which involves refining prompts, decoding interlayer representation and repairing. Facade, indoor scene and functional blocks like doors are supported in the generation. Experiments are conducted to evaluate the completeness and satisfaction of buildings generated via LLMs. It shows that LLMs hold significant potential for 3D building generation. Given appropriate prompts, LLMs can generate correct buildings in Minecraft with complete structures and incorporate specific building blocks such as windows and beds, meeting the specified requirements of human users.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑