Research map / Label-efficient and robust learning / Recognition and long-tail
Arbitrary Style Transfer With Deep Feature Reshuffle
Read the paperCheck 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 |
|---|---|---|---|
| Use reshuffle style loss instead of global Gram-matrix style loss (ablation) | Neural style transfer texture synthesis · 60 texture image pairs, layer 4 | global style loss | Helped |
| Use reshuffle style loss instead of local patch style loss (ablation) | Neural style transfer texture synthesis · 60 texture image pairs, layer 4 | local style loss | Helped |
| Allow higher maximum patch usage (ablation) | Deep Feature Reshuffle · 60 image pairs, layer 4 | global style loss | Hurt |
| Allow higher maximum patch usage (ablation) | Deep Feature Reshuffle · 60 image pairs, layer 4 | content loss | Helped |
| Increase patch size (ablation) | Deep Feature Reshuffle · Style transfer across layers | global style loss | Hurt |
| Increase patch size (ablation) | Deep Feature Reshuffle · Style transfer across layers | local style structure | Helped |
| Use reshuffle loss rather than either global or local style loss alone (ablation) | Neural style transfer · Texture synthesis | style quality | Mixed |
Closest papers by meaning
- Diversified Arbitrary Style Transfer via Deep Feature Perturbation (CVPR 2020 · similarity 0.90)
- Rethinking and Improving the Robustness of Image Style Transfer (CVPR 2021 · similarity 0.89)
- Style-Aware Normalized Loss for Improving Arbitrary Style Transfer (CVPR 2021 · similarity 0.89)
- MetaStyle: Three-Way Trade-off among Speed, Flexibility, and Quality in Neural Style Transfer (AAAI 2019 · similarity 0.89)
- AesFA: An Aesthetic Feature-Aware Arbitrary Neural Style Transfer (AAAI 2024 · 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.