Document Title

Interactive Layout Transfer

Document Information

  • Subject Area: Automation and Optimization in Graphical User Interface (GUI) Design
  • Keywords: User Interface Design, Computational Methods, MILP, Optimization, Style Transfer, Layout Redirection

Research Background and Problem

What problems or challenges did the authors identify?

  • The GUI design process must meet multiple constraints (e.g., design specifications, style guides, and design systems), and ensuring the final design is fully compliant poses a challenge.
  • Professional designers often need to repeatedly compare drafts with requirements to improve designs, which is time-consuming and can hinder creativity.
  • Searching for suitable designs among a vast number of examples to meet specific requirements is difficult and labor-intensive.

Why is this problem important?

  • In modern design workflows, quickly generating designs that adhere to specifications is crucial for improving efficiency and quality.
  • Large design systems and style guides require effective tools to ensure consistency and compliance; otherwise, design quality is prone to human error.

Research Motivation and Related Work

  • Current computational design methods are often applied to single design tasks, neglecting consistency with existing design examples and adherence to design specifications.
  • Existing layout redirection methods lack support for specification compliance or require strict mapping relationships between source and target layouts.
  • This paper aims to propose a method that can:
    1. Utilize design examples to achieve automated layout transfer while ensuring compliance with established specifications.
    2. Support interactive optimization of user interface layouts by designers.

Solution

What methods or solutions did the authors propose?

The authors proposed an interactive layout transfer method based on integer programming (IP), specifically:

  1. Layout Transfer Model:
    • Transform the source design layout into a target layout within a given design library while adhering to design specifications.
  2. Difference Calculation:
    • Introduce a Delta algorithm to measure differences between layouts, including size, position, and type differences.
  3. Design Library Search:
    • Automatically search for the most similar target layout within the given design library.
  4. Specification Constraint Model:
    • Use constraints (e.g., layout alignment, symmetry, padding) to ensure the design complies with specifications.

What are the innovative aspects of this solution?

  • Capable of handling complex scenarios where there is no clear one-to-one mapping between source and target layouts.
  • Proposes multi-dimensional "difference quantification" metrics to improve target matching and layout transfer.
  • Integrates optimization and design specifications to ensure the transferred layout is not only visually appealing but also fully compliant.
  • Supports interactive workflows for designers, such as providing optimization suggestions and control over element modifications.

What are the implementation steps and key technologies used?

  1. Delta Algorithm Calculation:
    • Calculate differences between two layouts (size, position, and type differences).
    • Model optimization using MILP (Mixed Integer Linear Programming).
  2. Design Library Search:
    • Compare the source layout with each layout in the design library, using the Delta algorithm to find the best match.
  3. Layout Optimization and Transfer:
    • Construct a new layout that combines the content of the source layout with the structure of the target layout.
    • Apply design specifications and aesthetic principles (e.g., alignment and padding) to generate the final layout.
  4. Interaction and Tool Implementation:
    • Integrate the method into the Sketch platform, allowing users to experience real-time, interactive design assistance.

Research Outcomes

What specific outcomes were achieved?

  • The method was successfully integrated into the Sketch design tool, enabling designers to avoid manual strict alignment.
  • Case studies and experiments demonstrated that the method significantly improves design productivity while ensuring high-quality and compliant designs.

How does it compare to existing solutions?

  • Higher Automation: Eliminates the need for designers to manually adjust mapping relationships between different layouts.
  • Specification Compliance: Not only transfers layout structures but also automates compliance with design systems and guidelines.
  • Enhanced Interactivity: Designers can flexibly select target structures and instantly view and modify generated results.

What were the experimental or evaluation results?

  1. Integration with Real-World Case Studies:
    • Tested within the UI design team of a large consumer electronics manufacturer, receiving positive feedback from designers.
    • The system significantly reduced manual workload in the design process.
  2. User Experiment (Perceived Quality):
    • Compared draft designs, manually optimized designs by designers, and system-generated optimized results.
    • System-generated designs showed significantly higher perceived quality than drafts and were comparable to manually optimized designs.
  3. Performance:
    • For layouts containing 20 to 30 elements, the main computational steps required minimal time (typically milliseconds to seconds), supporting interactive use.

Limitations and Future Directions

  • Limitations:

    • The current model only supports fixed layout tree structures and cannot dynamically expand.
    • The system relies on existing target layouts in the design library, and performance may be affected if suitable templates are lacking.
    • Has not been extensively tested for applications involving layouts with significantly different aspect ratios.
  • Future Directions:

    • Extend the model to support dynamic generation of target layouts.
    • Explore combining generative models to create new templates when the design library is insufficient.
    • Expand to cross-device and multi-platform design scenarios (e.g., responsive web design).

Quick Actions

Share

Share this page

ios_share

https://hci.top/en/papers/iui/58006/2021

AdRecommended

Learn AI Coding at CodeNow

open_in_newOpen DOI Link
DOI: https://doi.org/10.1145/3397481.3450652
At a Glance

Paper Snapshot

fact_check
dataset
Source
IUI
calendar_month
Year
2021
emoji_events
Award
No award tagged
group
Authors
6 authors
sell
Subtopics
Prototyping & User Testing
work
Professions
UI/UX Designers, HCI Researchers
article
Content Status
Full text indexed
hub
Related Papers
10 related papers