Research map / Image generation / Editing and style transfer
Diversified Arbitrary Style Transfer via Deep Feature Perturbation
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 deep feature perturbation to Li et al. WCT | Li et al. WCT · Style-transfer output pairs | Pixel distance: 0.000 -> 0.162 | Helped |
| Add deep feature perturbation to Li et al. WCT | Li et al. WCT · Style-transfer output pairs | LPIPS distance: 0.000 -> 0.431 | Helped |
| Add deep feature perturbation to Sheng et al. WCT | Sheng et al. WCT · Style-transfer output pairs | Pixel distance: 0.000 -> 0.102 | Helped |
| Add deep feature perturbation to Sheng et al. WCT | Sheng et al. WCT · Style-transfer output pairs | LPIPS distance: 0.000 -> 0.264 | Helped |
| Add deep feature perturbation to Li et al. WCT | Li et al. WCT · Style-transfer output pairs | Pixel distance: 0.000 -> 0.091 | Helped |
| Add deep feature perturbation to Li et al. WCT | Li et al. WCT · Style-transfer output pairs | LPIPS distance: 0.000 -> 0.203 | Helped |
| Use an orthogonal noise matrix instead of the original random noise matrix (ablation) | DFP-based style transfer · Arbitrary style transfer | stylization quality | Helped |
| Vary the orthogonal noise sampling distribution (ablation) | DFP-based style transfer · Arbitrary style transfer | visual output difference | No clear effect |
| Perturb shallower stylization levels instead of the deepest level (ablation) | Multi-level WCT style transfer · Multi-level stylization | stylization quality | Hurt |
| Perturb multiple stylization levels rather than only the deepest level (ablation) | Multi-level WCT style transfer · Multi-level stylization | stylization quality | Hurt |
| Increase the diversity weight λ (ablation) | DFP-based WCT methods · Arbitrary style transfer | stylization quality | Mixed |
| Increase diversity strength at larger stylization strength α (ablation) | Diversified WCT style transfer · Arbitrary style transfer | content structure preservation | Hurt |
Closest papers by meaning
- Universal Style Transfer via Feature Transforms (NeurIPS 2017 · similarity 0.91)
- Arbitrary Style Transfer With Deep Feature Reshuffle (CVPR 2018 · similarity 0.90)
- Learning Linear Transformations for Fast Image and Video Style Transfer (CVPR 2019 · similarity 0.88)
- OmniStyle: Filtering High Quality Style Transfer Data at Scale (CVPR 2025 · similarity 0.87)
- CLeaR: A Unified Framework for Resolving the Leakage–Degradation Dilemma in Style Transfer (NeurIPS 2026 · similarity 0.87)
- Puff-Net: Efficient Style Transfer with Pure Content and Style Feature Fusion Network (CVPR 2024 · similarity 0.86)
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.