pith. machine review for the scientific record. sign in

arxiv: 2604.11019 · v1 · submitted 2026-04-13 · 💻 cs.HC · cs.AI

Recognition: unknown

Brief2Design: A Multi-phased, Compositional Approach to Prompt-based Graphic Design

Kotaro Kikuchi , Nami Ogawa
Authors on Pith no claims yet

Pith reviewed 2026-05-10 16:05 UTC · model grok-4.3

classification 💻 cs.HC cs.AI
keywords graphic designprompt engineeringAI-assisted designdesign workflowshuman-AI collaborationrequirement extractioncompositional interfaces
0
0 comments X

The pith

A multi-phased compositional workflow for prompt-based graphic design increases prompt diversity and improves requirement handling over conversational baselines.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Through interviews with professional designers, the authors mapped how they turn abstract client briefs into visuals by first structuring requirements, then exploring separate elements, and finally recombining options. Brief2Design supports exactly this sequence with tools for requirement extraction and recommendation, element-specific generators for objects backgrounds text typography and composition, and free recombination of chosen pieces. In a within-subjects comparison against a standard chat interface, twelve designers produced more diverse prompts, rated the extraction and recommendation steps highly, yet spent more time generating images while achieving similar image variety. The work shows that decomposing the prompting task reduces fixation on premature complete outputs and clarifies vague goals at the direct cost of speed.

Core claim

Supporting the three-phase designer workflow of structuring ambiguous requirements, exploring alternatives at the level of individual elements, and recombining selected elements through a dedicated interface produces greater prompt diversity and stronger designer approval for requirement handling than unstructured conversational prompting with AI image generators.

What carries the argument

Brief2Design's three-phase interface that performs requirement extraction and recommendation, then element-level exploration for objects backgrounds text typography and composition, then flexible recombination of chosen alternatives.

If this is right

  • The structured workflow produces greater prompt diversity than direct conversation with an AI image model.
  • Designers assign high ratings to the requirement extraction and recommendation features.
  • Image generation requires more time under the multi-phased approach.
  • Final image diversity stays comparable to results from a conversational baseline.
  • AI graphic design tools must weigh gains in requirement clarity against losses in generation speed.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same phased decomposition could be tested in other prompt-driven creative tasks such as layout or illustration generation.
  • Adding automated suggestions for phase transitions might reduce the observed time penalty.
  • Longer-term use with repeated client feedback would reveal whether recombination supports iterative refinement better than one-shot generation.

Load-bearing premise

The workflow identified from interviews with six designers accurately reflects typical professional practice and the controlled study results with twelve participants will hold when designers apply the tool to live client projects.

What would settle it

A field study in which professional designers apply Brief2Design to real client briefs and generate no more diverse or better-matched prompts than with ordinary conversation would show the structured method brings no advantage.

Figures

Figures reproduced from arXiv: 2604.11019 by Kotaro Kikuchi, Nami Ogawa.

Figure 1
Figure 1. Figure 1: Brief2Design is an end-to-end system that supports designers in creating graphic designs from design briefs. The system [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Step 1 interface: Structuring free-form briefs into requirement cards. Users input their design brief (a) and trigger requirement [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Step 2 interface: Generating type-specific element cards with preview images. Users initiate element generation (a), and the [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Element card interface in Step 2. The left shows an element card with Edit, Delete, and Regenerate buttons. The right shows [PITH_FULL_IMAGE:figures/full_fig_p011_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Step 3 interface: Integrating selected elements into a final design. Users review their selected elements (a) from Step 2 and [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Baseline system interface: A conversational AI interface similar to ChatGPT. Users interact with the AI through text prompts [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Examples of submitted images from different participants (P2, P11, P7, and P6) under different task conditions (low-freedom [PITH_FULL_IMAGE:figures/full_fig_p016_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Image diversity comparison for generated and submitted images (ns: not significant). The circle markers indicate data points, [PITH_FULL_IMAGE:figures/full_fig_p017_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Prompt diversity comparison for generated and submitted images (ns: not significant, [PITH_FULL_IMAGE:figures/full_fig_p017_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Left: Scatter plot showing the relationship between prompt distance and image distance for the submitted images. The dotted [PITH_FULL_IMAGE:figures/full_fig_p018_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Results of the Creativity Support Index (CSI) questionnaire. [PITH_FULL_IMAGE:figures/full_fig_p019_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Rating proportions for each feature in the proposed system. [PITH_FULL_IMAGE:figures/full_fig_p020_12.png] view at source ↗
read the original abstract

Professional designers work from client briefs that specify goals and constraints but often lack concrete design details. Translating these abstract requirements into visual designs poses a central challenge, yet existing tools address specific aspects or induce fixation through complete outputs. Through interviews with six professional designers, we identified how designers address this challenge: first structuring ambiguous requirements, then exploring individual elements, and finally recombining alternatives. We developed Brief2Design, supporting this workflow through requirement extraction and recommendation, element-level exploration for objects, backgrounds, text, typography, and composition, and flexible recombination of selected elements. A within-subjects study with twelve designers compared Brief2Design against a conversational baseline. The structured approach increased prompt diversity and received high ratings for requirement extraction and recommendation, but required longer generation time and achieved comparable image diversity. These findings reveal that structured workflows benefit requirement clarification at the cost of efficiency, informing design trade-offs for AI-assisted graphic design tools.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The paper introduces Brief2Design, a multi-phased compositional system for prompt-based graphic design that structures the translation of ambiguous client briefs into visual outputs. Based on interviews with six professional designers, the workflow proceeds through requirement structuring/extraction/recommendation, element-level exploration (objects, backgrounds, text, typography, composition), and flexible recombination of alternatives. A within-subjects study with twelve designers compares Brief2Design to a conversational baseline, reporting increased prompt diversity, high subjective ratings for requirement extraction and recommendation, longer generation times, and comparable image diversity.

Significance. If the results hold under more rigorous validation, the work usefully surfaces efficiency-clarity trade-offs for AI-assisted graphic design tools and provides a concrete example of how phased, compositional prompting can mitigate fixation while supporting requirement clarification. The controlled within-subjects design and use of external diversity metrics against a baseline are strengths that could inform future tool-building in HCI.

major comments (3)
  1. [Evaluation] Evaluation section: the central claim that the structured workflow benefits requirement clarification rests on 12-participant within-subjects ratings for extraction and recommendation, yet no statistical details, effect sizes, order-effect analysis, or discussion of self-report biases are provided, leaving support for the claim only moderate.
  2. [Evaluation] Evaluation section: high ratings for requirement extraction lack any objective validation (e.g., expert scoring of how completely or accurately extracted requirements match the original client brief), so the reported benefit could reflect interface preference or demand characteristics rather than genuine clarification improvement.
  3. [Interviews and System Design] Interviews and System Design sections: the three-phase workflow is derived solely from interviews with six designers without further validation against observed practice or a larger sample; because the entire tool and study interpretation rest on this workflow, its representativeness is load-bearing.
minor comments (2)
  1. [Abstract] The abstract and results should explicitly define the prompt-diversity and image-diversity metrics and report the exact baseline values for direct comparison.
  2. [Evaluation] Clarify whether the within-subjects design counterbalanced task order and how generation-time measurements were normalized across conditions.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. We appreciate the recognition of the controlled within-subjects design and the identification of areas where greater rigor in reporting and discussion would strengthen the manuscript. We address each major comment below, indicating planned revisions where data or analysis can be added without new experiments.

read point-by-point responses

  1. Referee: [Evaluation] Evaluation section: the central claim that the structured workflow benefits requirement clarification rests on 12-participant within-subjects ratings for extraction and recommendation, yet no statistical details, effect sizes, order-effect analysis, or discussion of self-report biases are provided, leaving support for the claim only moderate.

    Authors: We agree that the current results section would benefit from more formal statistical reporting. Although the sample size is typical for within-subjects HCI studies and we prioritized direct participant comparisons and qualitative insights, we will add paired statistical tests (e.g., Wilcoxon signed-rank), effect sizes (e.g., rank-biserial correlation), and explicit analysis of order effects in the revised manuscript. We will also include a dedicated subsection discussing self-report biases and demand characteristics as limitations of the subjective measures. revision: yes

  2. Referee: [Evaluation] Evaluation section: high ratings for requirement extraction lack any objective validation (e.g., expert scoring of how completely or accurately extracted requirements match the original client brief), so the reported benefit could reflect interface preference or demand characteristics rather than genuine clarification improvement.

    Authors: This is a fair observation. Our study design emphasized designers' subjective perceptions of the tool's utility for requirement handling, which aligns with standard HCI evaluation practices for interactive systems. We did not collect objective expert ratings of requirement completeness or accuracy against the original briefs. In the revision we will expand the limitations and discussion sections to explicitly acknowledge this gap, note the possibility of interface preference or demand effects, and position objective validation as valuable future work. We cannot add such expert scoring without new data collection. revision: partial

  3. Referee: [Interviews and System Design] Interviews and System Design sections: the three-phase workflow is derived solely from interviews with six designers without further validation against observed practice or a larger sample; because the entire tool and study interpretation rest on this workflow, its representativeness is load-bearing.

    Authors: The three-phase workflow was derived through standard qualitative thematic analysis of interviews with six practicing designers, a sample size common in formative HCI research. We will revise the Interviews section to provide additional detail on participant selection criteria, interview protocol, and how themes were coded and validated internally. The subsequent 12-participant study provides empirical triangulation, as designers successfully applied the workflow in practice. We will add an explicit limitations paragraph noting that the workflow has not been validated against observational data or a larger sample and should be viewed as a starting point rather than a comprehensive model of all design practice. revision: partial

Circularity Check

0 steps flagged

No circularity in empirical derivation chain

full rationale

The paper derives its multi-phase workflow from interviews with six designers and evaluates it via a separate within-subjects study with twelve participants. Ratings for requirement extraction, prompt diversity metrics, and image diversity are computed from new participant data and generated outputs against an external baseline; none are defined in terms of the system itself or prior fitted values. No equations, self-citations, or uniqueness theorems appear in the provided text, and the central claims rest on independent empirical observations rather than definitional equivalence or fitted-input predictions.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an empirical HCI paper describing a new system and user study; it contains no mathematical derivations, fitted parameters, or postulated entities.

pith-pipeline@v0.9.0 · 5457 in / 1032 out tokens · 70654 ms · 2026-05-10T16:05:00.519343+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

46 extracted references · 13 canonical work pages

  1. [1]

    Shm Garanganao Almeda, J D Zamfirescu-Pereira, Kyu Won Kim, Pradeep Mani Rathnam, and Bjoern Hartmann. 2024. Prompting for discovery: Flexible sense-making for AI art-making with dreamsheets. InProceedings of the CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–17

    2024

  2. [2]

    Stephen Brade, Bryan Wang, Mauricio Sousa, Sageev Oore, and Tovi Grossman. 2023. Promptify: Text-to-image generation through interactive prompt exploration with large language models. InProceedings of the 36th Annual ACM Symposium on User Interface Software and Technology. ACM, New York, NY, USA

    2023

  3. [3]

    Yining Cao, Yiyi Huang, Anh Truong, Hijung Valentina Shin, and Haijun Xia. 2025. Compositional structures as substrates for human-AI co-creation environment: A design approach and A case study. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–25

    2025

  4. [4]

    Haoyu Chen, Xiaojie Xu, Wenbo Li, Jingjing Ren, Tian Ye, Songhua Liu, Ying-Cong Chen, Lei Zhu, and Xinchao Wang. 2025. POSTA: A Go-to Framework for Customized Artistic Poster Generation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 28694–28704

    2025

  5. [5]

    Erin Cherry and Celine Latulipe. 2014. Quantifying the creativity support of digital tools through the Creativity Support Index.ACM Trans. Comput. Hum. Interact.21, 4 (Aug. 2014), 1–25

    2014

  6. [6]

    Daeun Choi, Sumin Hong, Jeongeon Park, John Joon Young Chung, and Juho Kim. 2024. CreativeConnect: Supporting Reference Recombination for Graphic Design Ideation with Generative AI. InProceedings of the CHI Conference on Human Factors in Computing Systems (CHI ’24, Article 1055). Association for Computing Machinery, New York, NY, USA, 1–25

    2024

  7. [7]

    Steven P Dow, Alana Glassco, Jonathan Kass, Melissa Schwarz, Daniel L Schwartz, and Scott R Klemmer. 2010. Parallel prototyping leads to better design results, more divergence, and increased self-efficacy.ACM Trans. Comput. Hum. Interact.17, 4 (Dec. 2010), 1–24

    2010

  8. [8]

    Yingchaojie Feng, Xingbo Wang, Kam Kwai Wong, Sijia Wang, Yuhong Lu, Minfeng Zhu, Baicheng Wang, and Wei Chen. 2024. PromptMagician: Interactive prompt engineering for text-to-image creation.IEEE Trans. Vis. Comput. Graph.30, 1 (Jan. 2024), 295–305

    2024

  9. [9]

    Stephanie Fu, Netanel Tamir, Shobhita Sundaram, Lucy Chai, Richard Zhang, Tali Dekel, and Phillip Isola. 2023. DreamSim: Learning New Dimensions of Human Visual Similarity Using Synthetic Data. InAdvances in Neural Information Processing Systems, A. Oh, T. Naumann, A. Globerson, K. Saenko, M. Hardt, and S. Levine (Eds.), Vol. 36. 50742–50768

    2023

  10. [10]

    Frederic Gmeiner, Nicolai Marquardt, Michael Bentley, Hugo Romat, Michel Pahud, David Brown, Asta Roseway, Nikolas Martelaro, Kenneth Holstein, Ken Hinckley, and Nathalie Riche. 2025. Intent Tagging: Exploring Micro-Prompting Interactions for Supporting Granular Human-GenAI Co-Creation Workflows. InProceedings of the 2025 CHI Conference on Human Factors i...

    work page doi:10.1145/3706598.3713861 2025

  11. [11]

    Shunan Guo, Zhuochen Jin, Fuling Sun, Jingwen Li, Zhaorui Li, Yang Shi, and Nan Cao. 2021. Vinci: An intelligent graphic design system for generating advertising posters. InProceedings of the 2021 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA

    2021

  12. [12]

    Naoto Inoue, Kento Masui, Wataru Shimoda, and Kota Yamaguchi. 2024. OpenCOLE: Towards Reproducible Automatic Graphic Design Generation. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops. 8131–8135

    2024

  13. [13]

    Jansson and Steven M

    David G. Jansson and Steven M. Smith. 1991. Design Fixation.Design Studies12, 1 (1991), 3–11. doi:10.1016/0142-694X(91)90003-F

    work page doi:10.1016/0142-694x(91)90003-f 1991

  14. [14]

    Youngseung Jeon, Seungwan Jin, Patrick C Shih, and Kyungsik Han. 2021. FashionQ: An AI-driven creativity support tool for facilitating ideation in fashion design. InProceedings of the 2021 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA

    2021

  15. [15]

    Peidong Jia, Chenxuan Li, Yuhui Yuan, Zeyu Liu, Yichao Shen, Bohan Chen, Xingru Chen, Yinglin Zheng, Dong Chen, Ji Li, Xiaodong Xie, Shanghang Zhang, and Baining Guo. 2024. COLE: A Hierarchical Generation Framework for Multi-Layered and Editable Graphic Design. doi:10.48550/arXiv.2311.16974

    work page doi:10.48550/arxiv.2311.16974 2024

  16. [16]

    Wyn M Jones, Hedda Haugen Askland, et al. 2012. Design Briefs: Is There a Standard?. InProceedings of the International Conference on Engineering and Product Design Education. 115–120

    2012

  17. [17]

    Wenyuan Kong, Zhaoyun Jiang, Shizhao Sun, Zhuoning Guo, Weiwei Cui, Ting Liu, Jianguang Lou, and Dongmei Zhang. 2023. Aesthetics++: Refining Graphic Designs by Exploring Design Principles and Human Preference.IEEE Trans. Visual. Comput. Graphics29, 6 (2023), 3093–3104. doi:10.1109/TVCG.2022.3151617

    work page doi:10.1109/tvcg.2022.3151617 2023

  18. [18]

    LangChain. 2024. LangGraph

    2024

  19. [19]

    LangChain. 2025. Agent Chat UI. Brief2Design 25

    2025

  20. [20]

    Vivian Liu and Lydia B Chilton. 2022. Design guidelines for prompt engineering text-to-image generative models. InCHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA

    2022

  21. [21]

    Xuye Liu, Annie Sun, Pengcheng An, Tengfei Ma, and Jian Zhao. 2025. Influencer: Empowering everyday users in creating promotional posts via AI-infused exploration and customization. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–19

    2025

  22. [22]

    Peter O’Donovan, Aseem Agarwala, and Aaron Hertzmann. 2015. DesignScape: Design with Interactive Layout Suggestions. InProceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems. ACM, New York, NY, USA

    2015

  23. [23]

    OpenAI. 2024. OpenAI Platform — Text-Embedding-3-Large

    2024

  24. [24]

    OpenAI. 2025. Introducing Our Latest Image Generation Model in the API | OpenAI

    2025

  25. [25]

    Mayu Otani, Naoto Inoue, Kotaro Kikuchi, and Riku Togashi. 2024. LTSim: Layout Transportation-based Similarity Measure for Evaluating Layout Generation. doi:10.48550/arXiv.2407.12356

    work page doi:10.48550/arxiv.2407.12356 2024

  26. [26]

    Yifan Pu, Yiming Zhao, Zhicong Tang, Ruihong Yin, Haoxing Ye, Yuhui Yuan, Dong Chen, Jianmin Bao, Sirui Zhang, Yanbin Wang, Lin Liang, Lijuan Wang, Ji Li, Xiu Li, Zhouhui Lian, Gao Huang, and Baining Guo. 2025. ART: Anonymous Region Transformer for Variable Multi-Layer Transparent Image Generation. InProceedings of the IEEE/CVF Conference on Computer Visi...

    2025

  27. [27]

    Naina Raisinghani. 2025. Introducing Nano Banana Pro. https://blog.google/innovation-and-ai/products/nano-banana-pro/. (accessed 2026-01-19)

    2025

  28. [28]

    D Read, E Bohemia, et al. 2012. The Functions of the Design Brief. InProceedings of the International Design Conference. 1587–1596

    2012

  29. [29]

    Atefeh Shokrizadeh, Boniface Bahati Tadjuidje, Shivam Kumar, Sohan Kamble, and Jinghui Cheng. 2025. Dancing With Chains: Ideating Under Constraints With UIDEC in UI/UX Design. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. 1–23. doi:10.1145/ 3706598.3713785

    work page arXiv 2025

  30. [30]

    Auste Simkute, Lev Tankelevitch, Viktor Kewenig, Ava Elizabeth Scott, Abigail Sellen, and Sean Rintel. 2024. Ironies of Generative AI: Understanding and Mitigating Productivity Loss in Human-AI Interaction.International Journal of Human–Computer Interaction(2024), 1–22. doi:10.1080/10447318. 2024.2405782

    work page doi:10.1080/10447318 2024

  31. [31]

    Sangho Suh, Meng Chen, Bryan Min, Toby Jia-Jun Li, and Haijun Xia. 2024. Luminate: Structured generation and exploration of design space with large language models for human-AI co-creation. InProceedings of the CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–26

    2024

  32. [32]

    Sirui Tao, Ivan Liang, Cindy Peng, Zhiqing Wang, Srishti Palani, and Steven P Dow. 2025. DesignWeaver: Dimensional scaffolding for text-to-image product design. InProceedings of the 2025 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–26

    2025

  33. [33]

    Samangi Wadinambiarachchi, Ryan M Kelly, Saumya Pareek, Qiushi Zhou, and Eduardo Velloso. 2024. The effects of generative AI on design fixation and divergent thinking. InProceedings of the CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–18

    2024

  34. [34]

    Bryan Wang, Yuliang Li, Zhaoyang Lv, Haijun Xia, Yan Xu, and Raj Sodhi. 2024. LAVE: LLM-powered agent assistance and language augmentation for video editing. InProceedings of the 29th International Conference on Intelligent User Interfaces. ACM, New York, NY, USA

    2024

  35. [35]

    Heng Wang, Yotaro Shimose, and Shingo Takamatsu. 2025. BannerAgency: Advertising Banner Design with Multimodal LLM Agents. InProceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, Christos Christodoulopoulos, Tanmoy Chakraborty, Carolyn Rose, and Violet Peng (Eds.). 4304–4329. doi:10.18653/v1/2025.emnlp-main.214

    work page doi:10.18653/v1/2025.emnlp-main.214 2025

  36. [36]

    Zhijie Wang, Yuheng Huang, Da Song, Lei Ma, and Tianyi Zhang. 2024. PromptCharm: Text-to-image generation through multi-modal prompting and refinement. InProceedings of the CHI Conference on Human Factors in Computing Systems, Vol. 29. ACM, New York, NY, USA, 1–21

    2024

  37. [37]

    Tongshuang Wu, Michael Terry, and Carrie Jun Cai. 2022. AI chains: Transparent and controllable human-AI interaction by chaining large language model prompts. InCHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–22

    2022

  38. [38]

    Xuyong Yang, Tao Mei, Ying-Qing Xu, Yong Rui, and Shipeng Li. 2016. Automatic Generation of Visual-Textual Presentation Layout.ACM Trans. Multimedia Comput. Commun. Appl.12, 2, Article 33 (2016). doi:10.1145/2818709

    work page doi:10.1145/2818709 2016

  39. [39]

    Narasimhan, and Yuan Cao

    Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik R. Narasimhan, and Yuan Cao. 2023. ReAct: Synergizing Reasoning and Acting in Language Models. InThe Eleventh International Conference on Learning Representations

    2023

  40. [40]

    Weitao You, Yinyu Lu, Zirui Ma, Nan Li, Mingxu Zhou, Xue Zhao, Pei Chen, and Lingyun Sun. 2025. DesignManager: An Agent-Powered Copilot for Designers to Integrate AI Design Tools into Creative Workflows.ACM Trans. Graph.44, 4 (2025), 1–26. doi:10.1145/3730919

    work page doi:10.1145/3730919 2025

  41. [41]

    Youmans and Thomaz Arciszewski

    Robert J. Youmans and Thomaz Arciszewski. 2014. Design Fixation: Classifications and Modern Methods of Prevention.AI EDAM28, 2 (2014), 129–137. doi:10.1017/S0890060414000043

    work page doi:10.1017/s0890060414000043 2014

  42. [42]

    J D Zamfirescu-Pereira, Richmond Y Wong, Bjoern Hartmann, and Qian Yang. 2023. Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts. InProceedings of the 2023 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA, 1–21

    2023

  43. [43]

    Nanxuan Zhao, Nam Wook Kim, Laura Mariah Herman, Hanspeter Pfister, Rynson W H Lau, Jose Echevarria, and Zoya Bylinskii. 2020. ICONATE: Automatic compound icon generation and ideation. InProceedings of the 2020 CHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA

    2020

  44. [44]

    Jiayi Zhou, Renzhong Li, Junxiu Tang, Tan Tang, Haotian Li, Weiwei Cui, and Yingcai Wu. 2024. Understanding Nonlinear Collaboration between Human and AI Agents: A Co-design Framework for Creative Design. InProceedings of the CHI Conference on Human Factors in Computing Systems (CHI ’24, Article 170). Association for Computing Machinery, New York, NY, USA, 1–16

    2024

  45. [45]

    Tongyu Zhou, Jeff Huang, and Gromit Yeuk-Yin Chan. 2024. Epigraphics: Message-Driven Infographics Authoring. InProceedings of the CHI Conference on Human Factors in Computing Systems. 1–18. doi:10.1145/3613904.3642172 26 Kikuchi and Ogawa

    work page doi:10.1145/3613904.3642172 2024

  46. [46]

    {element_type}

    Xingxing Zou, Wen Zhang, and Nanxuan Zhao. 2025. From Fragment to One Piece: A Review on AI-Driven Graphic Design.Journal of Imaging11, 9 (2025), 289. doi:10.3390/jimaging11090289 A LLM Prompts This section presents the complete system prompts used in each LLM module of Brief2Design. Template variables are shown as{variable}. All LLM modules use OpenAI’s ...

    work page doi:10.3390/jimaging11090289 2025