arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2512.03335cs.CVcs.LG
收藏

逐步分层设计生成

Step-by-step Layered Design Generation

Faizan Farooq Khan, K J Joseph, Koustava Goswami, Mohamed Elhoseiny, Balaji Vasan Srinivasan

AI总结:

本文提出逐步分层设计生成方法,通过分层变化建模设计生成过程,引入新评估套件并验证其有效性。

摘要:

设计生成本质上是一个逐步的过程,设计师通过精心的修改逐步完善和提升他们的作品。尽管这一基本特征,现有方法主要将设计综合视为一步生成问题,显著低估了创造性过程的内在复杂性。为弥合这一差距,我们提出了一种新的问题设置,称为逐步分层设计生成,该设置要求机器学习模型生成一个遵循设计师一系列指令的设计。利用最近多模态大语言模型的进展,我们提出了SLEDGE:逐步分层设计生成器,以将每个设计的更新建模为对其前一状态的原子、分层变化,同时基于指令。为了补充我们的新问题设置,我们引入了一个新的评估套件,包括一个数据集和一个基准。我们的全面实验分析和与针对我们新设置的最新方法的比较证明了我们方法的有效性。我们希望我们的工作能引起对这一务实且未被充分探索的研究领域的关注。

英文摘要:

Design generation, in its essence, is a step-by-step process where designers progressively refine and enhance their work through careful modifications. Despite this fundamental characteristic, existing approaches mainly treat design synthesis as a single-step generation problem, significantly underestimating the inherent complexity of the creative process. To bridge this gap, we propose a novel problem setting called Step-by-Step Layered Design Generation, which tasks a machine learning model with generating a design that adheres to a sequence of instructions from a designer. Leveraging recent advancements in multi-modal LLMs, we propose SLEDGE: Step-by-step LayEred Design GEnerator to model each update to a design as an atomic, layered change over its previous state, while being grounded in the instruction. To complement our new problem setting, we introduce a new evaluation suite, including a dataset and a benchmark. Our exhaustive experimental analysis and comparison with state-of-the-art approaches tailored to our new setup demonstrate the efficacy of our approach. We hope our work will attract attention to this pragmatic and under-explored research area.

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。

↑