Towards Design Compositing

2026-04-16

AI Summary

Existing design generation methods typically assume input elements are stylistically consistent, struggling to handle heterogeneous components sourced from diverse origins with visual incoherence. To address this limitation, this work proposes GIST—a training-free, identity-preserving image synthesizer that harmonizes heterogeneous elements between layout prediction and typography generation, significantly enhancing overall visual coherence. GIST introduces, for the first time, identity-preserving stylized synthesis into the component-to-design generation pipeline, enabling seamless integration into existing design systems. Evaluations using multimodal large language models (LLaVA-OV and GPT-4V) demonstrate that GIST substantially outperforms naive collage baselines within the LaDeCo and Design-o-meter frameworks, achieving marked improvements in both aesthetic quality and visual consistency.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Graphic design creation involves harmoniously assembling multimodal components such as images, text, logos, and other visual assets collected from diverse sources, into a visually-appealing and cohesive design. Recent methods have largely focused on layout prediction or complementary element generation, while retaining input elements exactly, implicitly assuming that provided components are already stylistically harmonious. In practice, inputs often come from disparate sources and exhibit visual mismatch, making this assumption limiting. We argue that identity-preserving stylization and compositing of input elements is a critical missing ingredient for truly harmonized components-to-design pipelines. To this end, we propose GIST, a training-free, identity-preserving image compositor that sits between layout prediction and typography generation, and can be plugged into any existing components-to-design or design-refining pipeline without modification. We demonstrate this by integrating GIST with two substantially different existing methods, LaDeCo and Design-o-meter. GIST shows significant improvements in visual harmony and aesthetic quality across both pipelines, as validated by LLaVA-OV and GPT-4V on aspect-wise ratings and pairwise preference over naive pasting. Project Page: abhinav-mahajan10.github.io/GIST/.

Problem

• design compositing
• visual harmony
• identity-preserving stylization
• multimodal components
• graphic design

Innovation

• identity-preserving stylization
• visual harmony
• training-free compositing
• multimodal design
• design pipeline integration