Tell2Design: A Dataset for Language-Guided Floor Plan Generation
This work addresses the challenge of language-guided design generation for applications like architecture, but it is incremental as it builds on existing text-conditional image generation methods.
The authors tackled the problem of generating floor plans from natural language descriptions by introducing the Tell2Design dataset with over 80k designs and proposing a Sequence-to-Sequence baseline model, with human evaluations conducted to assess performance.
We consider the task of generating designs directly from natural language descriptions, and consider floor plan generation as the initial research area. Language conditional generative models have recently been very successful in generating high-quality artistic images. However, designs must satisfy different constraints that are not present in generating artistic images, particularly spatial and relational constraints. We make multiple contributions to initiate research on this task. First, we introduce a novel dataset, \textit{Tell2Design} (T2D), which contains more than $80k$ floor plan designs associated with natural language instructions. Second, we propose a Sequence-to-Sequence model that can serve as a strong baseline for future research. Third, we benchmark this task with several text-conditional image generation models. We conclude by conducting human evaluations on the generated samples and providing an analysis of human performance. We hope our contributions will propel the research on language-guided design generation forward.