Research map / Image generation / GANs
Spatial Fusion GAN for Image Synthesis
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 |
|---|---|---|---|
| Combine geometry and appearance synthesis instead of random alignment and appearance (ablation) | SF-GAN · Average across six text datasets | recognition accuracy: 47.8 -> 53.7 | Helped |
| Add the geometry synthesizer to the random-alignment baseline (ablation) | SF-GAN · Average across six text datasets | recognition accuracy: 47.8 -> 49.5 | Helped |
| Add the appearance synthesizer to the random-alignment baseline (ablation) | SF-GAN · Average across six text datasets | recognition accuracy: 47.8 -> 50.1 | Helped |
| Add appearance synthesis to geometry-synthesizer output (ablation) | SF-GAN · Average across six text datasets | recognition accuracy: 49.5 -> 53.7 | Helped |
| Add geometry synthesis to appearance synthesis with random alignments (ablation) | SF-GAN · Average across six text datasets | recognition accuracy: 50.1 -> 53.7 | Helped |
| Use SF-GAN-synthesized rather than Zhan-synthesized training images | Scene text recognition training · Average across six text datasets | recognition accuracy: 51.1 -> 53.7 | Helped |
| Use SF-GAN rather than ST-GAN to synthesize portrait accessories | Portrait wearing · Glass wearing | classified as real (%): 41.7 -> 62.0 | Helped |
Closest papers by meaning
- ST-GAN: Spatial Transformer Generative Adversarial Networks for Image Compositing (CVPR 2018 · similarity 0.90)
- Improving 3D-aware Image Synthesis with A Geometry-aware Discriminator (NeurIPS 2022 · similarity 0.88)
- SketchyGAN: Towards Diverse and Realistic Sketch to Image Synthesis (CVPR 2018 · similarity 0.87)
- Pi-GAN: Periodic Implicit Generative Adversarial Networks for 3D-Aware Image Synthesis (CVPR 2021 · similarity 0.87)
- Exploiting Spatial Dimensions of Latent in GAN for Real-Time Image Editing (CVPR 2021 · similarity 0.87)
- CGS-GAN: 3D Consistent Gaussian Splatting GANs for High Resolution Human Head Synthesis (NeurIPS 2025 · similarity 0.85)
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.