Research map / Detection and segmentation / Open-set and few-shot detection
Object-Centric Image Generation from Layouts
Read the paperCheck 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 |
|---|---|---|---|
| add scene-graph similarity supervision (SGSM) (ablation) | OC-GAN · COCO-Stuff, 64×64 | FID: 34.32 -> 29.57 | Helped |
| add scene-graph similarity supervision (SGSM) (ablation) | OC-GAN · COCO-Stuff, 64×64 | CA: 52.57 -> 60.27 | Helped |
| add the object discriminator (ablation) | OC-GAN · COCO-Stuff, 64×64 | CA: 48.03 -> 60.27 | Helped |
| add patch discriminators (ablation) | OC-GAN · COCO-Stuff, 64×64 | FID: 33.85 -> 29.57 | Helped |
| add patch discriminators (ablation) | OC-GAN · COCO-Stuff, 64×64 | CA: 62.48 -> 60.27 | Hurt |
| use two patch discriminators instead of one (ablation) | OC-GAN · COCO-Stuff, 64×64 | FID: 30.54 -> 29.57 | Helped |
| add bounding-box instance boundaries to conditioning (ablation) | OC-GAN · COCO-Stuff, 64×64 | FID: 30.12 -> 29.57 | Helped |
| add bounding-box instance boundaries to conditioning (ablation) | OC-GAN · COCO-Stuff, 64×64 | CA: 59.54 -> 60.27 | Helped |
| add perceptual loss (ablation) | OC-GAN · COCO-Stuff, 64×64 | FID: 31.14 -> 29.57 | Helped |
| add perceptual loss (ablation) | OC-GAN · COCO-Stuff, 64×64 | CA: 57.22 -> 60.27 | Helped |
| add SGSM when perceptual loss is absent (ablation) | OC-GAN without perceptual loss · COCO-Stuff, 64×64 | FID: 36.54 -> 31.14 | Helped |
| add SGSM when perceptual loss is absent (ablation) | OC-GAN without perceptual loss · COCO-Stuff, 64×64 | CA: 47.94 -> 57.22 | Helped |
Closest papers by meaning
- Exploiting Relationship for Complex-scene Image Generation (AAAI 2021 · similarity 0.92)
- Context-Aware Layout to Image Generation With Enhanced Object Appearance (CVPR 2021 · similarity 0.91)
- Generating Multiple Objects at Spatially Distinct Locations (ICLR 2019 · similarity 0.90)
- Generating Triples With Adversarial Networks for Scene Graph Construction (AAAI 2018 · similarity 0.89)
- Image Generation From Scene Graphs (CVPR 2018 · similarity 0.88)
- Image Generation From Layout (CVPR 2019 · similarity 0.88)
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.