Research map / Architectures and efficiency / Pruning and efficient tuning
Layering Virtual Try-On
Chun Feng, Bowei Chen, Shan Mengyi, Ira Kemelmacher-Shlizerman
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 stage 1 synthetic pre-training before layering fine-tuning (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | FID: 58.561 -> 48.957 | Improved |
| add stage 1 synthetic pre-training before layering fine-tuning (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | SSIM: 0.817 -> 0.843 | Improved |
| add stage 1 synthetic pre-training before layering fine-tuning (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | KID x10^3: 13.242 -> 1.463 | Improved |
| add stage 2 layering fine-tuning after stage 1 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | FID: 56.663 -> 48.957 | Improved |
| add stage 2 layering fine-tuning after stage 1 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | LPIPS: 0.183 -> 0.127 | Improved |
| add temporal reversal augmentation in stage 2 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | FID: 49.353 -> 48.957 | Improved |
| add temporal reversal augmentation in stage 2 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | KID x10^3: 3.003 -> 1.463 | Improved |
| add temporal reversal augmentation in stage 2 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | SSIM: 0.832 -> 0.843 | Improved |
| increase stage 1 data ratio relative to stage 2 (ablation) | Qwen-Image-Edit LoRA · LVTON test (532 pairs) | SSIM/LPIPS: gains plateau around 4.0x ratio | Improved |
| stage 1-only model evaluated zero-shot vs IDM-VTON (ablation) | Stage 1 model · VITON-HD | FID: 6.290 -> 5.235 | Improved |
| add stage 2 fine-tuning to stage 1 model (ablation) | Qwen-Image-Edit LoRA · VITON-HD | SSIM: 0.887 -> 0.928 | Improved |
| add stage 2 fine-tuning to stage 1 model (ablation) | Qwen-Image-Edit LoRA · DressCode | SSIM: 0.898 -> 0.948 | Improved |
Closest papers by meaning
- OpenVTON-Bench: A Large-Scale High-Resolution Benchmark for Controllable Virtual Try-On Evaluation (NeurIPS 2026 · similarity 0.90)
- MV-VTON: Multi-View Virtual Try-On with Diffusion Models (AAAI 2025 · similarity 0.89)
- M&M VTO: Multi-Garment Virtual Try-On and Editing (CVPR 2024 · similarity 0.89)
- BooW-VTON: Boosting In-the-Wild Virtual Try-On via Mask-Free Pseudo Data Training (CVPR 2025 · similarity 0.88)
- Cascaded Diffusion Models for Virtual Try-On: Improving Control and Resolution (AAAI 2025 · similarity 0.88)
- IMAGDressing-v1: Customizable Virtual Dressing (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.
