Research map / Image generation / Editing and style transfer
Learning Implicit Features with Flow-Infused Transformations for Realistic Virtual Try-On
Delong Zhang, Qiwei Huang, Yang Sun, Yuanliu Liu, Wei-Shi Zheng, Pengfei Xiong, Wei Zhang
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 Flow Guider (ablation) | FIA-VTON · VITON-HD | FID paired: 5.438 -> 4.686 | Improved |
| add Flow Guider (ablation) | FIA-VTON · VITON-HD | FID unpaired: 8.530 -> 7.914 | Improved |
| add Spatial Guider (ablation) | FIA-VTON · VITON-HD | FID unpaired: 8.645 -> 7.914 | Improved |
| add Spatial Guider (ablation) | FIA-VTON · VITON-HD | SSIM: 0.888 -> 0.913 | Improved |
| add Spatial Guider (ablation) | FIA-VTON · VITON-HD | PSNR: 24.330 -> 27.827 | Improved |
| use FashionCLIP instead of OpenCLIP as Spatial Guider (ablation) | FIA-VTON · VITON-HD | FID unpaired: 8.032 -> 7.914 | Improved |
| use FashionCLIP instead of OpenCLIP as Spatial Guider (ablation) | FIA-VTON · VITON-HD | SSIM: 0.914 -> 0.913 | No clear effect |
| use flow-infused attention instead of concatenated input (ablation) | FIA-VTON · VITON-HD | FID paired: 4.950 -> 4.686 | Improved |
| use flow-infused attention instead of plain cross-attention (ablation) | FIA-VTON · VITON-HD | FID paired: 5.28 -> 4.686 | Improved |
| use flow-infused attention instead of plain cross-attention (ablation) | FIA-VTON · VITON-HD | SSIM: 0.904 -> 0.913 | Improved |
| use bilinear instead of cubic downsampling for evaluation | FIA-VTON · VITON-HD | FID paired: 5.124 -> 4.686 | Mixed |
| FIA-VTON vs best warp-based D4-VTON | FIA-VTON · DressCode-Lower | FID unpaired: 14.86 -> 10.58 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- D4-VTON — D4-VTON: Dynamic Semantics Disentangling for Differential Diffusion based Virtual Try-On (ECCV 2024) · better on 8 of 8 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
- GP-VTON: Towards General Purpose Virtual Try-On via Collaborative Local-Flow Global-Parsing Learning (CVPR 2023 · similarity 0.91)
- ClothFormer: Taming Video Virtual Try-On in All Module (CVPR 2022 · similarity 0.90)
- Robust-MVTON: Learning Cross-Pose Feature Alignment and Fusion for Robust Multi-View Virtual Try-On (CVPR 2025 · similarity 0.88)
- MV-VTON: Multi-View Virtual Try-On with Diffusion Models (AAAI 2025 · similarity 0.88)
- Towards Squeezing-Averse Virtual Try-On via Sequential Deformation (AAAI 2024 · similarity 0.88)
- BooW-VTON: Boosting In-the-Wild Virtual Try-On via Mask-Free Pseudo Data Training (CVPR 2025 · similarity 0.87)
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.
