Research map / Image generation / GANs
Generative Visual Manipulation on the Natural Image Manifold
Jun-Yan Zhu, Philipp Krähenbühl, Eli Shechtman, Alexei A. Efros
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 |
|---|---|---|---|
| combine network prediction with optimization (hybrid) instead of optimization alone | hybrid projection · Shoes | reconstruction error: 0.155 -> 0.140 | Improved |
| combine network prediction with optimization (hybrid) instead of optimization alone | hybrid projection · Church Outdoor | reconstruction error: 0.319 -> 0.250 | Improved |
| combine network prediction with optimization (hybrid) instead of network alone | hybrid projection · Shirts | reconstruction error: 0.265 -> 0.184 | Improved |
| combine network prediction with optimization (hybrid) instead of network alone | hybrid projection · Shoes | reconstruction error: 0.210 -> 0.140 | Improved |
| train a class-specific generator instead of a cross-class model on shoes, handbags and shirts (ablation) | GAN projection · shoes, handbags, shirts | reconstruction error: ~10% lower | Improved |
| use a shoe model on shoes instead of on handbags (ablation) | class-specific model · shoes vs handbags | reconstruction error: 0.398 -> 0.140 | Improved |
| use a shoe model on shoes instead of on shirts (ablation) | class-specific model · shoes vs shirts | reconstruction error: 0.451 -> 0.140 | Improved |
| apply edit transfer to DCGAN samples (shape only) | DCGAN · AMT perception study | percent judged realistic: 14.3% -> 48.7% | Improved |
| add color transfer to shape-only edit transfer (ablation) | edit transfer · AMT perception study | percent judged realistic: 48.7% -> 25.9% | Hurt |
Closest papers by meaning
- Exploiting Spatial Dimensions of Latent in GAN for Real-Time Image Editing (CVPR 2021 · similarity 0.88)
- Drop the GAN: In Defense of Patches Nearest Neighbors As Single Image Generative Models (CVPR 2022 · similarity 0.87)
- Neural Photo Editing with Introspective Adversarial Networks (ICLR 2017 · similarity 0.87)
- On the "steerability" of generative adversarial networks (ICLR 2020 · similarity 0.87)
- Learning to Manipulate Artistic Images (AAAI 2024 · similarity 0.87)
- Pathways on the Image Manifold: Image Editing via Video Generation (CVPR 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.
