Research map / Image generation / Text-to-image diffusion
Property-Informed Diffusion-Based Text-to-Microstructure Generation
Bingxuan Dai, Hongsong Wang, Jie Gui
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add property-informed conditioning (ablation) | PropDiff-TMG · Geometries 2000 | Chamfer distance: 0.0651 -> 0.0395 | Improved |
| add property-informed conditioning (ablation) | PropDiff-TMG · Geometries 2000 | FID: 105.35 -> 70.81 | Improved |
| add contrastive text-structure alignment (ablation) | PropDiff-TMG · Geometries 2000 | FID: 264.63 -> 70.81 | Improved |
| add contrastive text-structure alignment (ablation) | PropDiff-TMG · Geometries 2000 | CLIP score: 0.5161 -> 0.6936 | Improved |
| add test-time reward-guided alignment (ablation) | PropDiff-TMG · Geometries 2000 | FID: 81.68 -> 70.81 | Improved |
| add test-time reward-guided alignment (ablation) | PropDiff-TMG · Geometries 2000 | CLIP score: 0.6078 -> 0.6936 | Improved |
| add discriminator reward to text-alignment reward (ablation) | PropDiff-TMG · Geometries 2000 | FID: 73.51 -> 70.81 | Improved |
| add discriminator reward to text-alignment reward (ablation) | PropDiff-TMG · Geometries 2000 | CLIP score: 0.7038 -> 0.6936 | Hurt |
| add reward normalization (ablation) | PropDiff-TMG · Geometries 2000 | CLIP score: 0.7189 -> 0.6936 | Hurt |
| add reward normalization (ablation) | PropDiff-TMG · Geometries 2000 | FID: 77.52 -> 70.81 | Improved |
| add property-informed conditioning (ablation) | PropDiff-TMG · GenText-Microstruct | CLIP score: 0.5210 -> 0.6463 | Improved |
| add test-time reward-guided alignment (ablation) | PropDiff-TMG · GenText-Microstruct | CLIP score: 0.5164 -> 0.6463 | Improved |
Closest papers by meaning
- Periodic Materials Generation using Text-Guided Joint Diffusion Model (ICLR 2025 · similarity 0.88)
- MatExpert: Decomposing Materials Discovery By Mimicking Human Experts (ICLR 2025 · similarity 0.86)
- UniMate: A Unified Model for Mechanical Metamaterial Generation, Property Prediction, and Condition Confirmation (ICML 2025 · similarity 0.86)
- Fine-Tuned Language Models Generate Stable Inorganic Materials as Text (ICLR 2024 · similarity 0.85)
- FlowLLM: Flow Matching for Material Generation with Large Language Models as Base Distributions (NeurIPS 2024 · similarity 0.85)
- Generating Symmetric Materials using Latent Flow Matching (NeurIPS 2026 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
