2022年的後疫情時代，ChatGPT的發布引起了全球對於生成式人工智慧（AI）的關注。生成式AI大幅降低了AI的入門門檻，並擴展了使用範疇，各產業紛紛嘗試將其應用於專業領域，建築產業也面臨人工智慧驅動數位轉型的挑戰。
本論文開發了一套結合擴增實境（AR）與生成式AI的「即時風格化設計渲染」系統，命名為DreamLite，基於智慧型行動裝置，提供便攜、直觀、快速設計產出的工具，旨在探討XR與AI技術整合應用在移動式建築設計流程中的潛力。
DreamLite乃是透過AR放置家具模型或虛擬積木，截圖後傳送至運算主機端的生成式AI進行圖像生成，並回傳至行動裝置與雲端儲存。在極短的時間內產出基於現實卻又風格多樣的提案渲染圖，協助即時設計討論與檢討。即時風格化設計渲染系統可以應用在三種使用情境：1. 基地現場的設計討論（實際尺寸模擬）、2. 會議或設計課中的討論（縮尺模型模擬）及3. 參與式設計的意見整合。
本研究進一步進行使用需求分析，邀請了4位潛在使用者試用，分別來自建築背景與非建築背景。另外也發放了問卷並附上示範影片，共蒐集20份有效回饋。根據反饋優化系統設計，並針對相關議題進行反思。結果顯示，DreamLite系統有助於靈感發想與設計溝通。
AI生成設計對建築師風格獨特性與事務所價值定位的影響值得深入研究。生成式AI將設計流程中的渲染呈現提早至初期提案階段，而AR協助將生成內容落實。若進一步整合混合實境（MR）、建築資訊模型（BIM）、多模態（Multimodal）格式等技術，有望發展為更全面的系統性服務。

This thesis developed a "real-time stylized design rendering" system, named DreamLite, which combines augmented reality (AR) and generative AI. Operating on smart mobile devices, it provides a portable, intuitive, and rapid design tool that aims to explore the potential of integrating XR and AI technologies in mobile architectural design processes.

This research conducts a user needs analysis by inviting four potential users from both architectural and non-architectural backgrounds to test the system. Additionally, questionnaires were distributed with a demonstration video, resulting in 20 valid responses. Based on the feedback, the system was optimized and related issues were reflected upon. The results indicate that the DreamLite system facilitates inspiration generation and design communication. Generative AI advances rendering to the early proposal stage, while AR helps realize the generated content. This study shows the potential of integrating different technologies into a design tool to improve the current architectural design process.

書面資料
Borji, A. (2022). Generated Faces in the Wild: Quantitative Comparison of Stable Diffusion, Midjourney and DALL-E 2. ArXiv, abs/2210.00586.
Broll, W., Lindt, I., Herbst, I., Ohlenburg, J., Braun, A.-K., & Wetzel, R. (2008). Toward Next-Gen Mobile AR Games. IEEE Computer Graphics and Applications, 28(4), 40-48. https://doi.org/10.1109/mcg.2008.85
Caudell, T. P., & Mizell, D. W. (1992, 7-10 Jan. 1992). Augmented reality: an application of heads-up display technology to manual manufacturing processes. Proceedings of the Twenty-Fifth Hawaii International Conference on System Sciences,
Gavrilov, E. (2019 ). Magnetizing Floor Plan Generator. https://toolbox.decodingspaces.net/magnetizing-floor-plan-generator/
Hegazy, M., & Saleh, A. (2023). Evolution of AI role in architectural design: between parametric exploration and machine hallucination. MSA Engineering Journal, 2(2), 262-288. https://doi.org/10.21608/msaeng.2023.291873
Kalay, Y. E. (2004). Architecture's New Media: Principles, Theories, and Methods of Computer-aided Design. MIT Press. https://books.google.com.tw/books?id=BDboJQJvUq8C
Kolarevic, B. (2004). Architecture in the Digital Age: Design and Manufacturing. Taylor & Francis. https://books.google.com.tw/books?id=L-p4AgAAQBAJ
Kumar, S., & Rai, S. (2012). Survey on Transport Layer Protocols: TCP & UDP. International Journal of Computer Applications, 46, 20-25.
Lee, K. (2012). Augmented Reality in Education and Training. TechTrends, 56. https://doi.org/10.1007/s11528-012-0559-3
Milgram, P., Takemura, H., Utsumi, A., & Kishino, F. (1994). Augmented reality: A class of displays on the reality-virtuality continuum. Telemanipulator and Telepresence Technologies, 2351. https://doi.org/10.1117/12.197321
Park, S., Bokijonov, S., & Choi, Y. (2021). Review of Microsoft HoloLens Applications over the Past Five Years. Applied Sciences, 11, 7259. https://doi.org/10.3390/app11167259
Radford, A., Kim, J. W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., & Clark, J. (2021). Learning transferable visual models from natural language supervision. International conference on machine learning,
Ramesh, A., Dhariwal, P., Nichol, A., Chu, C., & Chen, M. (2022). Hierarchical text-conditional image generation with clip latents. arXiv preprint arXiv:2204.06125, 1(2), 3.
Roumeliotis, K. I., & Tselikas, N. D. (2023). ChatGPT and Open-AI Models: A Preliminary Review. Future Internet, 15(6), 192. https://doi.org/10.3390/fi15060192
Sawicki, J., Ganzha, M., & Paprzycki, M. (2023). The State of the Art of Natural Language Processing—A Systematic Automated Review of NLP Literature Using NLP Techniques. Data Intelligence, 5(3), 707-749. https://doi.org/10.1162/dint_a_00213
Yang, L., Zhang, Z., Song, Y., Hong, S., Xu, R., Zhao, Y., Zhang, W., Cui, B., & Yang, M.-H. (2022). Diffusion Models: A Comprehensive Survey of Methods and Applications. ACM Computing Surveys, 56(4), 1-39. https://doi.org/10.1145/3626235
Yeh, Z.-F., Lai, S.-Y., Liu, D.-E., Hsu, C.-C., Chang, F.-Y., Tsai, M.-Z., & Lin, R.-H. (2024, 2024). Footprints of Travel: AIoT and AR Enhanced Tourist Gaming Experience in Unmanned Cultural Sites.
Zhu, J.-Y., Park, T., Isola, P., & Efros, A. A. (2017). Unpaired image-to-image translation using cycle-consistent adversarial networks. Proceedings of the IEEE international conference on computer vision,
網路資料
Bruner, J., & Deshpande, A. (2018). Generative Adversarial Networks for Beginners. https://github.com/jonbruner/generative-adversarial-networks/blob/master/gan-notebook.ipynb
Campo, M. d. (2023). Artificial Intelligence and Architecture: Matias del Campo https://www.youtube.com/watch?v=pyBMASbjlyg&ab_channel=ComputationalDesignDetroit
Google. (2009). Sky Map. https://play.google.com/store/apps/details?id=com.google.android.stardroid&pcampaignid=web_share
Lewis, T. (2013). Medical app uses augmented reality to enhance patient education. https://www.imedicalapps.com/2013/07/medical-app-augmented-reality-patient-education/
Niantic. (2024). Pokémon GO. https://pokemongolive.com/zh_hant/
NVIDIA. (2023). Text2Materials Demo | NVIDIA Research at #SIGGRAPH2023. https://youtu.be/nmTEuPIriLY?si=2KHxu7uSsqhpsx0n
NVIDIA. (2024). NVIDIA Omniverse. https://www.nvidia.com/zh-tw/omniverse/
Stark, J. (2020). Autodesk Extends the Power of Generative Design to Architecture, Engineering and Construction Industries. https://adsknews.autodesk.com/en/news/generative-design-revit/