Research map / Image generation / Autoregressive and latent generation
Enhancing Photorealism Enhancement
Stephan R. Richter, Hassan Abu Alhaija, Vladlen Koltun
Read the paperCheck 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 matching-patch sampling instead of uniform sampling with crop 400 (ablation) | Ours · GTA to Cityscapes | sKVD 2-2: 72.94973 -> 11.11894 | Improved |
| Use matching-patch sampling instead of uniform sampling with crop 196 (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 2.629832 -> 6.128937 | Hurt |
| Use smaller uniform crops (400 to 196) (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 43.80217 -> 2.629832 | Improved |
| Add full G-buffers vs no G-buffers (ablation) | Ours · GTA to Cityscapes | sKVD 3-3: 34.7575 -> 15.45489 | Improved |
| Add full G-buffers vs no G-buffers (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 5.843171 -> 6.128937 | Hurt |
| Use full G-buffer set instead of VIPER subset (ablation) | Ours · GTA to Cityscapes | sKVD 2-2: 15.70158 -> 11.11894 | Improved |
| Use RAD modules instead of concatenation (ablation) | Ours · GTA to Cityscapes | sKVD 3-3: 90.2788 -> 15.45489 | Improved |
| Use RAD modules instead of SPADE modules (ablation) | Ours · GTA to Cityscapes | sKVD 3-3: 296.1047 -> 15.45489 | Improved |
| Use perceptual discriminator instead of PatchGAN (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 45.45944 -> 6.128937 | Improved |
| Add segmentation projection to discriminator (ablation) | Ours · GTA to Cityscapes | sKVD 3-3: 19.0776 -> 15.45489 | Improved |
| Add segmentation projection to discriminator (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 3.26578 -> 6.128937 | Hurt |
| Add adaptive backpropagation to discriminator (ablation) | Ours · GTA to Cityscapes | sKVD 1-2: 8.089039 -> 6.128937 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TSIT — TSIT: A Simple and Versatile Framework for Image-to-Image Translation (ECCV 2020) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Deep CG2Real: Synthetic-to-Real Translation via Image Disentanglement (ICCV 2019 · similarity 0.88)
- Image Visual Realism: From Human Perception to Machine Computation (TPAMI 2017 · similarity 0.87)
- Neural Photo Editing with Introspective Adversarial Networks (ICLR 2017 · similarity 0.86)
- Generative Photography: Scene-Consistent Camera Control for Realistic Text-to-Image Synthesis (CVPR 2025 · similarity 0.86)
- CAD: Photorealistic 3D Generation via Adversarial Distillation (CVPR 2024 · similarity 0.85)
- Real3D: The Curious Case of Neural Scene Degeneration (AAAI 2024 · 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.
