Research map / Image generation / Editing and style transfer
StyTr2: Image Style Transfer With Transformers
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 |
|---|---|---|---|
| Use a transformer-based style-transfer architecture instead of CNN-based encoder-decoder methods | StyTr2 · Repeated stylization | content structure preservation | Helped |
| Use content-aware, scale-invariant positional encoding instead of sinusoidal positional encoding (ablation) | StyTr2 · Different input resolutions and repetitive-pattern content | visual quality | Helped |
| Use content-aware positional encoding instead of sinusoidal positional encoding (ablation) | StyTr2 · Repetitive-pattern content images | consistency of stylized regions | Helped |
| Use a transformer architecture rather than a CNN-based encoder-decoder pipeline | StyTr2 · MS-COCO and WikiArt; 800 stylized images | content loss (Lc): 1.91 vs. 1.97 | Helped |
| Use StyTr2 rather than SANet | Image style transfer · MS-COCO and WikiArt; 800 stylized images | style loss (Ls): 1.47 vs. 1.18 | Hurt |
| Use StyTr2 rather than IEST | Image style transfer · MS-COCO and WikiArt; 800 stylized images | content loss (Lc): 1.97 -> 1.91 | Helped |
| Use StyTr2 rather than AdaIN | Image style transfer · MS-COCO and WikiArt; 800 stylized images | content loss (Lc): 2.34 -> 1.91 | Helped |
| Use StyTr2 rather than AdaIN | Image style transfer · MS-COCO and WikiArt; 800 stylized images | style loss (Ls): 1.91 -> 1.47 | Helped |
Closest papers by meaning
- A Content Transformation Block for Image Style Transfer (CVPR 2019 · similarity 0.89)
- Master: Meta Style Transformer for Controllable Zero-Shot and Few-Shot Artistic Style Transfer (CVPR 2023 · similarity 0.88)
- Puff-Net: Efficient Style Transfer with Pure Content and Style Feature Fusion Network (CVPR 2024 · similarity 0.88)
- S2WAT: Image Style Transfer via Hierarchical Vision Transformer Using Strips Window Attention (AAAI 2024 · similarity 0.87)
- Image Style Transfer Using Convolutional Neural Networks (CVPR 2016 · similarity 0.87)
- MicroAST: Towards Super-fast Ultra-Resolution Arbitrary Style Transfer (AAAI 2023 · 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.