Research map / Image generation / Editing and style transfer
FACT: Fused Attention for Clothing Transfer with Generative Adversarial Networks
Yicheng Zhang, Lei Li, Li Song, Rong Xie, Wenjun 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 soft attention to base model (ablation) | Base Model · DeepFashion | FID: 36.03 -> 34.71 | Improved |
| add self-attention on top of soft attention (ablation) | Base Model + SoA · DeepFashion | FID: 34.71 -> 32.67 | Improved |
| add stylized channel-wise attention (ablation) | Base Model + SoA + SeA · DeepFashion | FID: 32.67 -> 30.54 | Improved |
| add all three fused attention modules (ablation) | Base Model · DeepFashion | FID: 36.03 -> 30.54 | Improved |
| FACT vs FashionGAN (ablation) | FashionGAN · DeepFashion | FID: 35.18 -> 30.54 | Improved |
| add stylized channel-wise attention (ablation) | Base Model + SoA + SeA · DeepFashion attribute prediction | Has Long Pants accuracy: 89.6 -> 91.0 | Improved |
| add stylized channel-wise attention (ablation) | Base Model + SoA + SeA · DeepFashion attribute prediction | Has Shorts accuracy: 90.0 -> 90.2 | Improved |
| add soft and self-attention (ablation) | Base Model · DeepFashion attribute prediction | Has Jeans accuracy: 80.6 -> 82.7 | Improved |
| fused attention base model vs no attention (ablation) | Base Model · DeepFashion attribute prediction | Has T-Shirt accuracy: 61.4 -> 64.1 | Improved |
| FACT vs FashionGAN | FashionGAN · User study | Rank 1 share: 14.58% -> 51.76% | Improved |
| base model without attention vs FashionGAN | FashionGAN · DeepFashion | FID: 35.18 -> 36.03 | Hurt |
Later papers that compared with this method
3 later accepted papers report FacT as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Parameter Efficient Adaptation for Image Restoration with Heterogeneous Mixture-of-Experts (NeurIPS 2024 · better on 1 of 1 shared result)
- EDGE: Editable Dance Generation From Music (CVPR 2023 · better on 4 of 4 shared results)
- A Brand New Dance Partner: Music-Conditioned Pluralistic Dancing Controlled by Multiple Dance Genres (CVPR 2022 · better on 3 of 3 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
- FashionR2R: Texture-preserving Rendered-to-Real Image Translation with Diffusion Models (NeurIPS 2024 · similarity 0.86)
- AnyDressing: Customizable Multi-Garment Virtual Dressing via Latent Diffusion Models (CVPR 2025 · similarity 0.85)
- GarFast: Realistic and Fast Garment Transfer with a Simplified Parser-Free Approach (AAAI 2025 · similarity 0.85)
- DreamFit: Garment-Centric Human Generation via a Lightweight Anything-Dressing Encoder (AAAI 2025 · similarity 0.85)
- Attention Distillation: A Unified Approach to Visual Characteristics Transfer (CVPR 2025 · similarity 0.85)
- IMAGDressing-v1: Customizable Virtual Dressing (AAAI 2025 · similarity 0.85)
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.
