Research map / Image generation / Autoregressive and latent generation
Image Warping for Image-to-Image Translation
Shen Zheng, Anurag Ghosh, Gaurav Parmar, Srinivasa G. Narasimhan
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 |
|---|---|---|---|
| Add warp-unwarp (ablation) | pix2pix-Turbo relighting · VITON-HD human relighting | user score avg: 3.56 -> 4.33 | Improved |
| Add warp-unwarp (ablation) | pix2pix-Turbo relighting · StreetTryOn human relighting | user score avg: 3.57 -> 4.23 | Improved |
| Add Background Prompt Substitution to data pipeline (ablation) | pix2pix-Turbo relighting · VITON-HD human relighting | user score avg: 3.67 -> 4.33 | Improved |
| Add Background Prompt Substitution to data pipeline (ablation) | pix2pix-Turbo relighting · StreetTryOn human relighting | user score avg: 3.55 -> 4.23 | Improved |
| Add warp-unwarp (ablation) | pix2pix-Turbo relighting · ROADWork golden sunlight | user score avg: 4.25 -> 4.54 | Improved |
| Add warp-unwarp (ablation) | pix2pix-Turbo relighting · ROADWork foggy | user score avg: 4.26 -> 4.51 | Improved |
| Add warp-unwarp with GT boxes | CycleGAN-Turbo · BDD100K Day->Night | FID: 19.2 -> 17.5 | Improved |
| Add warp-unwarp with GT boxes | CycleGAN-Turbo · Cityscapes->ACDC Foggy | FID: 308.8 -> 169.2 | Improved |
| Use detector boxes instead of GT boxes for saliency (ablation) | CycleGAN-Turbo + Warp · BDD100K Day->Night | FID: 17.5 -> 17.7 | No clear effect |
| Add warp-unwarp with GT boxes | CycleGAN-Turbo · DarkZurich->Cityscapes | DINO-Struct: 2.22 -> 2.75 | Hurt |
| Add warp-unwarp to pix2pix-Turbo | pix2pix-Turbo · VITON 768x1024 | latency: 0.892 -> 0.898 s (+0.006) | No clear effect |
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.
- DreamLight — DreamLight: Towards Harmonious and Consistent Image Relighting (NeurIPS 2025) · better on 4 of 4 shared results
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
- UniRelight: Learning Joint Decomposition and Synthesis for Video Relighting (NeurIPS 2025 · similarity 0.87)
- FashionR2R: Texture-preserving Rendered-to-Real Image Translation with Diffusion Models (NeurIPS 2024 · similarity 0.86)
- FRESCO: Spatial-Temporal Correspondence for Zero-Shot Video Translation (CVPR 2024 · similarity 0.86)
- UniLumos: Fast and Unified Image and Video Relighting with Physics-Plausible Feedback (NeurIPS 2025 · similarity 0.85)
- MultiDiff: Consistent Novel View Synthesis from a Single Image (CVPR 2024 · similarity 0.85)
- Constructive Distortion: Improving MLLMs with Attention-Guided Image Warping (ICLR 2026 · 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.
