Property-Informed Diffusion-Based Text-to-Microstructure Generation

2026-06-06

AI Summary

This work addresses the challenge in existing inverse design methods for 3D metamaterial microstructures, which often struggle to simultaneously achieve sufficient design diversity and physical feasibility. To overcome this limitation, the authors propose a diffusion-based text-to-microstructure generation framework that jointly leverages semantic and physical property information from natural language prompts to guide the diffusion process. The approach introduces two key innovations: contrastive text-structure alignment during training and reward-guided alignment at inference time, enabling the generation of diverse microstructures that are both semantically consistent with input descriptions and physically realizable. Experimental results demonstrate that the method efficiently produces a wide range of 3D microstructures across multiple material classes, satisfying target physical constraints while maintaining clear semantic correspondence, thereby highlighting its strong potential for language-driven, interactive materials design.

Abstract

Designing 3D metamaterial microstructures that meet the intended functions remains a major challenge, as it typically requires domain expertise, iterative simulations, and extensive manual tuning. Existing work on inverse design that automatically generates microstructures based on desired target properties often suffers from limited design diversity and faces challenges in ensuring the physical feasibility of the generated structures. To address this issue, a property-informed diffusion-based network is proposed that enables the generation of 3D microstructures directly from textual descriptions. Unlike traditional property conditioning methods, our approach leverages rich guidance in terms of semantics and physical properties in the text input to support diverse structure synthesis. To enforce consistency between the generated structures and the target textual prompts, a dual alignment strategy is adopted, including contrastive text-structure alignment and test-time reward-guided alignment. Experimental results show that the model is capable of generating semantically meaningful and physically plausible structures across a wide range of material categories. Our approach has good potential for interactive microstructure design and opens up new directions for combining language-based interfaces with inverse material discovery. Code is available at: https://github.com/hongsong-wang/PropDiff-TMG

Problem

• inverse design
• 3D microstructures
• property conditioning
• physical feasibility
• design diversity

Innovation

• property-informed diffusion
• text-to-microstructure generation
• dual alignment strategy
• inverse design
• 3D metamaterials