Research map / Image generation / Editing and style transfer
Domain-Aware Universal Style Transfer
Kibeom Hong, Seogkyu Jeon, Huan Yang, Jianlong Fu, Hyeran Byun
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 domain-aware skip connections (ablation) | DSTN · photo-realistic style transfer (qualitative) | structure preservation | Improved |
| pass full encoded features through skip connections instead of only high-frequency components (ablation) | DSTN · artistic style transfer (qualitative) | stylization degree | Hurt |
| pass and transform high-frequency components only (ablation) | DSTN · photo and artistic references (qualitative) | content preservation vs stylization | Improved |
| use contextual similarity instead of pixel-wise L2 for reconstruction (ablation) | DSTN decoder · artistic style transfer (qualitative) | texture detail / blur | Improved |
| use DSTN versus Avatar-Net/AAMS/Colla-Distil in user study | DSTN · User study, artistic methods | Preservation preference (photo refs): 0.76% -> 92.42% | Improved |
| use DSTN versus WCT2 in user study | DSTN · User study, photo-realistic methods | Stylization preference (artistic refs): 6.01% -> 80.62% | Improved |
| use DSTN versus LST in user study | DSTN · User study, multi-domain methods | Preservation preference (photo refs): 43.94% -> 29.17% | Hurt |
| use DSTN versus LST in user study | DSTN · User study, multi-domain methods | Stylization preference (artistic refs): 27.71% -> 62.41% | Improved |
| vary domainness alpha continuously between 0 and 1 (ablation) | DSTN · interpolation (qualitative) | stylization level | Improved |
Closest papers by meaning
- Universal Style Transfer via Feature Transforms (NeurIPS 2017 · similarity 0.90)
- Learning Linear Transformations for Fast Image and Video Style Transfer (CVPR 2019 · similarity 0.89)
- Style Transfer by Relaxed Optimal Transport and Self-Similarity (CVPR 2019 · similarity 0.89)
- A Style-Aware Content Loss for Real-time HD Style Transfer (ECCV 2018 · similarity 0.89)
- Style Injection in Diffusion: A Training-free Approach for Adapting Large-scale Diffusion Models for Style Transfer (CVPR 2024 · similarity 0.88)
- SigStyle: Signature Style Transfer via Personalized Text-to-Image Models (AAAI 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.
