Research map / Image generation / Text-to-image diffusion
PromptDresser: Improving the Quality and Controllability of Virtual Try-On via Generative Textual Prompt and Prompt-aware Mask
Jeongho Kim, Hoiyeong Jin, Sunghyun Park, Jaegul Choo
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 prompt-aware mask generation (PMG) at inference (ablation) | PromptDresser · VITON-HD | FID: 8.98 -> 8.54 | Improved |
| add PMG at inference (ablation) | PromptDresser · VITON-HD | SSIM: 0.8411 -> 0.8686 | Improved |
| add PMG at inference (ablation) | PromptDresser · VITON-HD | LPIPS: 0.1485 -> 0.1119 | Improved |
| use structured attribute captions instead of holistic description (ablation) | PromptDresser · VITON-HD | FID: 9.21 -> 8.54 | Improved |
| use structured attribute captions instead of holistic description (ablation) | PromptDresser · VITON-HD | KID: 1.18 -> 0.67 | Improved |
| use GPT-4o captions instead of LLaVA (ablation) | PromptDresser · VITON-HD | FID: 8.85 -> 8.54 | Improved |
| add generative text prompt to IDM-VTON | IDM-VTON · VITON-HD | FID: 9.20 -> 9.15 | Improved |
| add generative text prompt to IDM-VTON | IDM-VTON · VITON-HD | LPIPS: 0.1023 -> 0.1025 | No clear effect |
| add DensePose channels (13-channel input) | PromptDresser · VITON-HD | LPIPS: 0.1119 -> 0.0967 | Improved |
| add DensePose channels | PromptDresser · VITON-HD | KID: 0.67 -> 1.16 | Hurt |
| add DensePose channels | PromptDresser · Text alignment 'tight fit' | Accuracy: 66.98% -> 44.09% | Hurt |
| lower sigma from 0.8 to 0.3 for PMG early stopping (ablation) | PromptDresser · VITON-HD | KID: 0.68 -> 0.65 | No clear effect |
Later papers that compared with this method
2 later accepted papers report PromptDresser as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- High-Fidelity Virtual Try-On beyond Paired Data Scarcity via Diffusion-based Cycle-Consistent Learning (CVPR 2026 · better on 3 of 3 shared results)
- PROMO: Promptable Outfitting for Efficient High-Fidelity Virtual Try-On (CVPR 2026 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- IMAGDressing-v1: Customizable Virtual Dressing (AAAI 2025 · similarity 0.91)
- M&M VTO: Multi-Garment Virtual Try-On and Editing (CVPR 2024 · similarity 0.90)
- Enhancing Virtual Try-On with Synthetic Pairs and Error-Aware Noise Scheduling (CVPR 2025 · similarity 0.89)
- TexFit: Text-Driven Fashion Image Editing with Diffusion Models (AAAI 2024 · similarity 0.89)
- Cascaded Diffusion Models for Virtual Try-On: Improving Control and Resolution (AAAI 2025 · similarity 0.88)
- DreamFit: Garment-Centric Human Generation via a Lightweight Anything-Dressing Encoder (AAAI 2025 · similarity 0.88)
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.
