Research map / Image generation / Editing and style transfer
StyleBank: An Explicit Representation for Neural Image Style Transfer
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 |
|---|---|---|---|
| Train a new style by updating only its StyleBank filter bank while fixing the encoder and decoder | StyleBank · Incremental style training | training time speedup: 20~40 times | Helped |
| Train a new style by updating only its StyleBank filter bank while fixing the encoder and decoder | StyleBank · Incremental style training | training time: around 8 minutes with about $1,000$ iterations | Helped |
| Use $(7,7)$ rather than $(3,3)$ StyleBank kernels (ablation) | StyleBank · StyleBank kernel-size visualization | style element size | Helped |
| Reduce the maximum channel number from 128 to 32 (ablation) | StyleBank · StyleBank model-size comparison | stylization quality | Hurt |
| Reduce the maximum channel number from 128 to 64 (ablation) | StyleBank · StyleBank model-size comparison | stylization quality | Mixed |
| Add the auto-encoder reconstruction branch during training (ablation) | StyleBank · Auto-encoder reconstruction | reconstruction quality | Helped |
| Remove residual blocks from the encoder-decoder architecture (ablation) | StyleBank · Style transfer | performance | No clear effect |
| Use the StyleBank representation to share one auto-encoder across multiple styles | StyleBank · Style transfer comparisons | region-specific stylization | Helped |
Closest papers by meaning
- Image Style Transfer Using Convolutional Neural Networks (CVPR 2016 · similarity 0.89)
- A Flexible Convolutional Solver for Fast Style Transfers (CVPR 2019 · similarity 0.89)
- Real-Time Neural Style Transfer for Videos (CVPR 2017 · similarity 0.89)
- Learning Linear Transformations for Fast Image and Video Style Transfer (CVPR 2019 · similarity 0.88)
- ArtBank: Artistic Style Transfer with Pre-trained Diffusion Model and Implicit Style Prompt Bank (AAAI 2024 · similarity 0.88)
- OmniStyle: Filtering High Quality Style Transfer Data at Scale (CVPR 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.