Research map / Image generation / Editing and style transfer
Single Stage Virtual Try-On via Deformable Attention Flows
Shuai Bai, Huiling Zhou, Zhikang Li, Chang Zhou, Hongxia Yang
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 |
|---|---|---|---|
| increase deformable attention sampling points K from 1 (conventional single flow) to 6 (ablation) | SDAFN · VITON (paired) | PSNR: 22.43 -> 26.48 | Improved |
| increase deformable attention sampling points K from 1 to 6 (ablation) | SDAFN · VITON (paired) | SSIM: 0.833 -> 0.888 | Improved |
| increase K from 6 to 8 (ablation) | SDAFN · VITON (paired) | PSNR: 26.48 -> 25.35 | Hurt |
| add cascade flow estimation to the baseline (ablation) | Baseline · VITON (paired) | PSNR: 20.88 -> 21.83 | Improved |
| add shallow encoder and decoder (ablation) | Baseline + Cascade · VITON (paired) | PSNR: 21.83 -> 22.43 | Improved |
| add the deformable attention flow network (DAFN) (ablation) | Baseline + Cascade + Shallow En. · VITON (paired) | PSNR: 22.43 -> 26.48 | Improved |
| add the deformable attention flow network (DAFN) (ablation) | Baseline + Cascade + Shallow En. · VITON (paired) | SSIM: 0.833 -> 0.888 | Improved |
| compare SDAFN with strongest prior ZFlow in paired setting | ZFlow · VITON (paired) | FID: 15.17 -> 10.97 | Improved |
| train SDAFN+ with SDAFN predictions as person input (parser-free, like PF-APN) instead of SDAFN | SDAFN · VITON (unpaired) | FID: 12.05 -> 9.46 | Improved |
| SDAFN+ vs PF-APN on MPV | PF-APN · MPV (unpaired) | FID: 6.429 -> 5.805 | Improved |
| flow interpolation instead of image interpolation for higher resolution (512x384) without retraining (ablation) | SDAFN · VITON-HD images | visual texture clarity | Improved |
Closest papers by meaning
- Style-Based Global Appearance Flow for Virtual Try-On (CVPR 2022 · similarity 0.90)
- IMAGDressing-v1: Customizable Virtual Dressing (AAAI 2025 · similarity 0.88)
- MV-VTON: Multi-View Virtual Try-On with Diffusion Models (AAAI 2025 · similarity 0.87)
- Virtual Try-On: Real-Time Interactive Hybrid Network with High-Fidelity (AAAI 2024 · similarity 0.87)
- Incorporating Visual Correspondence into Diffusion Model for Virtual Try-On (ICLR 2025 · similarity 0.87)
- Towards Squeezing-Averse Virtual Try-On via Sequential Deformation (AAAI 2024 · 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.
