Research map / Image generation / Editing and style transfer
Look Closer To Supervise Better: One-Shot Font Generation via Component-Based Discriminator
Yuxin Kong, Canjie Luo, Weihong Ma, Qiyuan Zhu, Shenggao Zhu, Nicholas Yuan, Lianwen Jin
Check 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 |
|---|---|---|---|
| Use component-level supervision instead of pixel-level L1 supervision (ablation) | CG-GAN · Chinese font generation, unseen styles and unseen contents | FID: 51.44 -> 17.94 | Improved |
| Use component-level rather than character-level supervision (ablation) | CG-GAN · Chinese font generation, unseen styles and unseen contents | FID: 33.21 -> 17.94 | Improved |
| Add structure retention loss (ablation) | CG-GAN CAM · Chinese font generation, unseen styles and unseen contents | FID: 49.09 -> 22.33 | Improved |
| Add style matching loss to structure retention loss (ablation) | CG-GAN CAM · Chinese font generation, unseen styles and unseen contents | FID: 22.33 -> 18.67 | Improved |
| Add component realism loss (ablation) | CG-GAN CAM · Chinese font generation, unseen styles and unseen contents | FID: 18.67 -> 17.94 | Improved |
| Remove the AdaIN style latent vector (ablation) | CG-GAN · Chinese font generation, unseen styles and unseen contents | FID: 17.94 -> 18.73 | Hurt |
| Remove the U-Net discriminator decoder D_dec (ablation) | CG-GAN · Chinese font generation, unseen styles and unseen contents | FID: 6.54 -> 7.14; 17.94 -> 19.07 | Mixed |
| Remove the U-Net discriminator decoder D_dec (ablation) | CG-GAN · IAM writer-relevant handwriting generation | FID: 102.18 -> 101.48; 110.07 -> 111.29; 104.81 -> 102.67; 113.01 -> 112.77 | Mixed |
Closest papers by meaning
- Multi-Content GAN for Few-Shot Font Style Transfer (CVPR 2018 · similarity 0.90)
- Multiple Heads Are Better Than One: Few-Shot Font Generation With Multiple Localized Experts (ICCV 2021 · similarity 0.90)
- Generate Like Experts: Multi-Stage Font Generation by Incorporating Font Transfer Process into Diffusion Models (CVPR 2024 · similarity 0.89)
- VQ-FONT: Few-Shot Font Generation with Structure-Aware Enhancement and Quantization (AAAI 2024 · similarity 0.87)
- VecGlypher: Unified Vector Glyph Generation with Language Models (CVPR 2026 · similarity 0.86)
- FontDiffuser: One-Shot Font Generation via Denoising Diffusion with Multi-Scale Content Aggregation and Style Contrastive Learning (AAAI 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.
