Treffer: 基于大语言模型和实时视觉反馈的 城市设计教学系统开发及应用.
Weitere Informationen
To address challenges in urban design education such as insufficient spatial imagination, linear collaboration, and delayed feedback, this study integrated physical scale models (1:500 scale), ESP32-CAM camera array vision capture technology, large language model requirement analysis capabilities, and diffusion-based image generation methods to construct a "teacher-student" interactive design system. Multimodal inputs were jointly encoded, producing high-fidelity street scenes within an average of 10 seconds; physical operations synchronously drove the digital model, and operation logs triggered structured diagnostics. Empirical results showed that the system improved the accuracy of conveying design intent, reduced the number of reworks, and increased collaboration efficiency. Additionally, it could rapidly deliver regulatory checks and optimization suggestions, thereby enhancing students’ decision-making confidence and course satisfaction. The results demonstrated that the system enabled concrete expression of requirements, dynamic collaboration, and real-time feedback, providing a viable technical path for a cognitive-constructivist paradigm in urban design education. [ABSTRACT FROM AUTHOR]
为突破城市设计教学中空间想象不足、协作线性及反馈滞后等难题, 本研究通过整合实体白模 (1∶500 比 例) 、ESP32-CAM 相机阵列视觉采集技术、大语言模型需求解析能力与扩散式图像生成方法, 构建"师生"交互设计 系统。多模态输入经联合编码, 平均 10 s 内输出高保真街景; 物理操作同步驱动数字模型, 操作日志触发结构化诊 断。实证结果表明, 该系统能够提升设计意图传达的准确性, 减少返工次数并提高协作效率。同时, 可在短时间内推 送规范检测与优化建议, 从而增强学生的决策信心与课程满意度。结果证明, 该系统实现了需求具象化、协作动态 化、反馈实时化, 为认知建构范式的城市设计教学提供了可行技术路径。 [ABSTRACT FROM AUTHOR]
Copyright of Journal of Nanchang University (Natural Science) is the property of Journal of Nanchang University (Natural Science) Editorial Office and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)