Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

DesignFromX: Empowering Consumer-Driven Design Space Exploration through Feature Composition of Referenced Products

Runlin Duan, Chenfei Zhu, Yuzhao Chen, Yichen Hu, Jingyu Shi, Karthik Ramani

TL;DR

This work tackles the challenge of enabling non-expert consumers to explore and shape product designs by extracting and composing design features from visual references using GenAI. It introduces DesignFromX, a three-module workflow (reference component query, feature analysis, design composition) that generates concept images and 3D models while supporting an iterative, incremental design process. A formative study with novices and a two-phase user study (N=24) show that DesignFromX yields broader feature exploration, higher engagement, and more favorable perceived design outcomes than a baseline system, with some trade-offs in image quality and controllability due to GenAI limitations. The results suggest a promising direction for end-user participation in design and point to future work integrating controllable generation, finer-grained feature control, and professional-design workflows.

Abstract

Industrial products are designed to satisfy the needs of consumers. The rise of generative artificial intelligence (GenAI) enables consumers to easily modify a product by prompting a generative model, opening up opportunities to incorporate consumers in exploring the product design space. However, consumers often struggle to articulate their preferred product features due to their unfamiliarity with terminology and their limited understanding of the structure of product features. We present DesignFromX, a system that empowers consumer-driven design space exploration by helping consumers to design a product based on their preferences. Leveraging an effective GenAI-based framework, the system allows users to easily identify design features from product images and compose those features to generate conceptual images and 3D models of a new product. A user study with 24 participants demonstrates that DesignFromX lowers the barriers and frustration for consumer-driven design space explorations by enhancing both engagement and enjoyment for the participants.

DesignFromX: Empowering Consumer-Driven Design Space Exploration through Feature Composition of Referenced Products

TL;DR

This work tackles the challenge of enabling non-expert consumers to explore and shape product designs by extracting and composing design features from visual references using GenAI. It introduces DesignFromX, a three-module workflow (reference component query, feature analysis, design composition) that generates concept images and 3D models while supporting an iterative, incremental design process. A formative study with novices and a two-phase user study (N=24) show that DesignFromX yields broader feature exploration, higher engagement, and more favorable perceived design outcomes than a baseline system, with some trade-offs in image quality and controllability due to GenAI limitations. The results suggest a promising direction for end-user participation in design and point to future work integrating controllable generation, finer-grained feature control, and professional-design workflows.

Abstract

Industrial products are designed to satisfy the needs of consumers. The rise of generative artificial intelligence (GenAI) enables consumers to easily modify a product by prompting a generative model, opening up opportunities to incorporate consumers in exploring the product design space. However, consumers often struggle to articulate their preferred product features due to their unfamiliarity with terminology and their limited understanding of the structure of product features. We present DesignFromX, a system that empowers consumer-driven design space exploration by helping consumers to design a product based on their preferences. Leveraging an effective GenAI-based framework, the system allows users to easily identify design features from product images and compose those features to generate conceptual images and 3D models of a new product. A user study with 24 participants demonstrates that DesignFromX lowers the barriers and frustration for consumer-driven design space explorations by enhancing both engagement and enjoyment for the participants.
Paper Structure (60 sections, 13 figures, 4 tables)

This paper contains 60 sections, 13 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Visual and Textual Reference for Design Space Exploration
  4. GenAI-based Design Support Tools
  5. Zero-shot prompting on Large Language Model
  6. Design Rationale
  7. Formative Study
  8. Study Insights
  9. Study Insight 1: Barriers to Decomposing Visual References
  10. Study Insight 2: Challenges in Articulating Aesthetic and Functional Design Preferences
  11. Study Insight 3: Facilitation of Exploration by Examples
  12. Study Insight 4: Preference of Increment Design
  13. Design Goals
  14. DesignFromX
  15. System Workflow
  16. ...and 45 more sections

Figures (13)

  • Figure 1: An overview of DesignFromX workflow. (A) Users draw keypoints on the reference image to query a component; (B) The system segments and analyzes the queried component; (C) A feature analysis of the queried component is provided by the system, where users can choose the features they admire; (D) Users indicate the components or regions they want to modify on the initial design; (E) The system integrates selected design features into initial designs to compose new designs; (F) New designs are visualized for exploration.
  • Figure 2: DesignFromX system user interface. A) A major design canvas showcasing the updated design; B) A reference image canvas that enables interactions, like querying component, with reference images; C) an interactive pop-up design feature table, triggered by querying a component on the reference image; D) An image gallery to explore the design space of a product; E) A communication portal to search for reference images in various styles and from various sources; F) A history tracker that documents the changes made during design iterations, and allow for undo modifications.
  • Figure 3: 3D model generation from single design image
  • Figure 4: Demographics of the participants in the user study phase 2
  • Figure 5: Workflow of the user study Phase 2
  • ...and 8 more figures