許多自動化方法已運用於設計與藝術的生成上，包括機器學習、風格移轉，或利用設計原理所訂定的生成規則。然而，沒有研究系統地比較各種自動化設計方法，這使得未來的研究人員和設計師很難考慮其他可行方案。本論文旨為比較各種可行的自動化設計方法，為簡單起見，研究將限縮於Facebook廣告的生成。通過對先前視覺設計方法和設計理論的回顧，我們以設計理念、算法族群和操作細節進行分類，提出了自動視覺設計方法的分類學。最終，我們實施了分類學所提之方法，以更深入地了解這些技術的差異以及這些差異對採用公司的影響。

Several approaches have been used to attempt automated design and art, including machine learning, constraint rules based on design principles, or even style transfer. However, there are no studies that systematically compare automated design approaches, which makes it hard for future researchers and designers to consider their alternatives. We seek to compare and contrast different feasible approaches for automating design. We limited our implementation to a subset of Facebook advertisements for sake of simplicity. From our review of previous visual design approaches and design theories, we propose a taxonomy of approaches for automated visual design, distinguished by design philosophy, algorithm families, and operational details. We implemented various approaches in this taxonomy to get a stronger qualitative sense of how these techniques differ and the implications of these differences to firms.

Abstract 2
Acknowledgment 4
1. Introduction 6
2. Visual design of online advertisements 8
2.1 Overview of online advertisements 8
2.2 Previous automated visual design approaches 10
2.3 Challenge of automated online advertisement design 12
3. Alternative approaches for automated design 13
3.1 Design decision philosophy 14
3.2 Algorithm family 15
3.3 Continuous vs. Heuristics implementation 17
4. Study Design and Data 19
4.1. Overview of Study 19
4.2. Data Collection 20
4.3 Data Annotation 20
4.4 Data structure 23
5. Implementation of automated ad design 24
5.1 Descriptive approach 24
5.2 Predictive learning 27
5.3 Style transfer 32
6. Analysis and Results 33
6.1 Descriptive approach results 34
6.2 Predictive learning results 36
6.3 Style transfer results 38
6.4 Overview of results 40
7. Discussion 42
7.1 Contributions 42
7.2 Limitations and future work 43
8. Reference 43

Arnheim, R. (1965). Art and visual perception: A psychology of the creative eye. Univ of California Press.
Arnold, L., Rebecchi, S., Chevallier, S., & Paugam-Moisy, H. (2011, April). An Introduction to Deep Learning. In European Symposium on Artificial Neural Networks (ESANN), Bruges (Belgium).
Colormind. (n.d.). Generating color palettes. Retrieved from http://colormind.io/blog/
Evans, P., & Thomas, M. A. (2012). Exploring the elements of design. Cengage Learning.
Forsey, J. (2016). The aesthetics of design. Oxford University Press.
Ganesan, A., & Oates, T. (2017). Fashioning with networks: Neural style transfer to design clothes. arXiv preprint arXiv:1707.09899.
Gatarski, R. (2002). Breed better banners: Design automation through on-line interaction. Journal of Interactive Marketing, 16(1), 2-13.
Itti, L., Koch, C., & Niebur, E. (1998). A model of saliency-based visual attention for rapid scene analysis. IEEE Transactions on Pattern Analysis & Machine Intelligence, (11), 1254-1259.
Izumi, Ko. (2018, September 20). 「AIのバナー広告自動生成」のタネ明かし!?. Retrieved from https://dentsu-ho.com/articles/6247?fbclid=IwAR11cpf0IRBbFZ2x7ab8269UZiCsDRoPg2rKon7tbmPSAOmpuqt8WlKIAyo
Jahanian, A., Liu, J., Tretter, D. R., Lin, Q., Damera-Venkata, N., O'Brien-Strain, E., ... & Allebach, J. P. (2012, February). Automatic design of magazine covers. In Imaging and Printing in a Web 2.0 World III(Vol. 8302, p. 83020N). International Society for Optics and Photonics.
Jahanian, A. (2016). Quantifying aesthetics of visual design applied to automatic design. Springer.
Koffka, K. (1922). Perception: an introduction to the Gestalt-Theorie. Psychological Bulletin, 19(10), 531.
Krotov, V., & Silva, L. (2018). Legality and ethics of web scraping.
Kwon, O. H., & Ma, K. L. (2019). A Deep Generative Model for Graph Layout. arXiv preprint arXiv:1904.12225.
Lok, S., & Feiner, S. (2001). A survey of automated layout techniques for information presentations. Proceedings of SmartGraphics, 2001, 61-68.
Lok, S., Feiner, S., & Ngai, G. (2004, January). Evaluation of visual balance for automated layout. In Proceedings of the 9th international conference on Intelligent user interfaces (pp. 101-108). ACM.
Looka. (n.d.). What is Looka and How Does it Work?. Retrieved from https://looka.com/blog/what-is-looka-and-how-does-it-work/
Montabone, S., & Soto, A. (2010). Human detection using a mobile platform and novel features derived from a visual saliency mechanism. Image and Vision Computing, 28(3), 391-402.
Nicolas, P. R. (2015). Scala for machine learning. Packt Publishing Ltd.
Norman, D. (2013). The design of everyday things: Revised and expanded edition. Basic books.
Rosenholtz, R., Dorai, A., & Freeman, R. (2011). Do predictions of visual perception aid design?. ACM Transactions on Applied Perception (TAP), 8(2), 12.
Shmueli, G., Bruce, P. C., Yahav, I., Patel, N. R., & Lichtendahl Jr, K. C. (2017). Data mining for business analytics: concepts, techniques, and applications in R. John Wiley & Sons.
Sternberg, R. J., & Sternberg, K. (2016). Cognitive psychology. Nelson Education.
Tezign. (2017, April 28). 電商 CONNECT with AI，雙 11 期間阿里的人工智慧「魯班」設計了 1.7 億個 BANNER. Retrieved from https://mp.weixin.qq.com/s?__biz=MjM5NTMzNzg3NA==&mid=2650605092&idx=1&sn=4de8d9c85380104d1756a6ce48033244&chksm=bef3a5d989842ccfce1a08f240d95853bba5233d9bc0c297500974dcd1fead3e437f8775b49f&mpshare=1&scene=1&srcid=0424dbUtEd9SjOxDCeFpePRP&key=953bef6a6e503c01ff1bc30a5d008af23843690df4c2ab2d604126a96619e7712eae12f15a1fd4993037368144c015911b956a17c92752ed9c5161e3db9a20ec74a8a7b61b5172397982ee98a3de3642&ascene=0&uin=MTUxNjQyMjU0Mg%3D%3D&devicetype=iMac+MacBookPro11%2C1+OSX+OSX+10.12.4+build(16E195)&version=12020110&nettype=WIFI&fontScale=100&pass_ticket=BrR8ObqD8c3D4%2B7av%2Fd5X7%2BDGNiaEIhe7PjfAcVmqq%2Fu0ZNe07V6lGqt9BnNF9Zl
Tomita, K. (2015). Principles and elements of visual design: A review of the literature on visual design studies of instructional materials. Educational Studies (Institute for Educational Research and Service, International Christian University), 57, 167-174.
Yang, X., Mei, T., Xu, Y. Q., Rui, Y., & Li, S. (2016). Automatic generation of visual-textual presentation layout. ACM Transactions on Multimedia Computing, Communications, and Applications (TOMM), 12(2), 33.
Zhang, Y., Hu, K., Ren, P., Yang, C., Xu, W., & Hua, X. S. (2017, October). Layout Style Modeling for Automating Banner Design. In Proceedings of the on Thematic Workshops of ACM Multimedia 2017 (pp. 451-459). ACM.