Research map / Image generation / Autoregressive and latent generation
SketchyCOCO: Image Generation From Freehand Scene Sketches
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 shared attribute vector to connect generated edge maps and images | EdgeGAN · SketchyCOCO object-level | FID: 141.5 -> 87.6 | Helped |
| Generate foreground objects before conditioning background synthesis on those objects | EdgeGAN · SketchyCOCO scene-level | FID (local): 123.0 -> 112.0 | Helped |
| Add the joint discriminator D_J (ablation) | EdgeGAN · SketchyCOCO object-level | FID: 95.63 -> 87.59 | Helped |
| Add the joint discriminator D_J (ablation) | EdgeGAN · SketchyCOCO object-level | Accuracy: 0.8361 -> 0.8866 | Helped |
| Add the joint discriminator D_J (ablation) | EdgeGAN · SketchyCOCO object-level | Shape Similarity: 2.457e+04 -> 2.294e+04 | Helped |
| Add the image discriminator D_I (ablation) | EdgeGAN · SketchyCOCO object-level | FID: 110.17 -> 87.59 | Helped |
| Add the image discriminator D_I (ablation) | EdgeGAN · SketchyCOCO object-level | Accuracy: 0.6964 -> 0.8866 | Helped |
| Add the image discriminator D_I (ablation) | EdgeGAN · SketchyCOCO object-level | Shape Similarity: 2.331e+04 -> 2.294e+04 | Helped |
| Add the edge discriminator D_E (ablation) | EdgeGAN · SketchyCOCO object-level | FID: 91.12 -> 87.59 | Helped |
| Add the edge discriminator D_E (ablation) | EdgeGAN · SketchyCOCO object-level | Accuracy: 0.8204 -> 0.8866 | Helped |
| Add the edge discriminator D_E (ablation) | EdgeGAN · SketchyCOCO object-level | Shape Similarity: 2.341e+04 -> 2.294e+04 | Helped |
| Use WGAN-GP instead of DCGAN (ablation) | EdgeGAN · SketchyCOCO object-level | FID: 108.86 -> 87.59 | Helped |
Closest papers by meaning
- SketchyGAN: Towards Diverse and Realistic Sketch to Image Synthesis (CVPR 2018 · similarity 0.89)
- SketchingReality: From Freehand Scene Sketches to Photorealistic Images (ICLR 2026 · similarity 0.88)
- Visual Object Networks: Image Generation with Disentangled 3D Representations (NeurIPS 2018 · similarity 0.88)
- TextureGAN: Controlling Deep Image Synthesis With Texture Patches (CVPR 2018 · similarity 0.87)
- BachGAN: High-Resolution Image Synthesis From Salient Object Layout (CVPR 2020 · similarity 0.87)
- A Layer-Based Sequential Framework for Scene Generation with GANs (AAAI 2019 · 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.