Research map / Image generation / Autoregressive and latent generation
Generating Diverse Structure for Image Inpainting With Hierarchical VQ-VAE
Jialun Peng, Dong Liu, Songcen Xu, Houqiang Li
Check 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 full attention rather than cross attention on learned features (ablation) | Texture generator · CelebA-HQ | FID: 11.26 -> 11.55 | Improved |
| Use structural features instead of learned features for attention (ablation) | Texture generator · CelebA-HQ | FID: 11.26 -> 10.12 | Improved |
| Use full attention on structural features (ablation) | Texture generator · CelebA-HQ | FID: 11.55 -> 9.670 | Improved |
| Add structural feature loss (ablation) | Texture generator · CelebA-HQ | SSIM: 0.8581 -> 0.8589 | Improved |
| Add textural feature loss (ablation) | Texture generator · CelebA-HQ | FID: 11.02 -> 10.34 | Improved |
| Add both structural and textural feature losses (ablation) | Texture generator · CelebA-HQ | FID: 11.02 -> 9.670 | Improved |
| Replace proposed feature losses with perceptual and style losses (ablation) | Texture generator · CelebA-HQ | IS: 3.467 -> 3.388 | Hurt |
| Use Ours rather than PIC | Multiple-solution inpainting · CelebA-HQ test | FID: 11.70 -> 9.784 | Improved |
| Use Ours rather than UCTGAN | Multiple-solution inpainting · CelebA-HQ test | FID: 11.74 -> 9.784 | Improved |
| Use Ours rather than PIC | Multiple-solution inpainting · CelebA-HQ test | LPIPS diversity: 0.029 vs 0.029 | 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.
- UCTGAN — UCTGAN: Diverse Image Inpainting Based on Unsupervised Cross-Space Translation (CVPR 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
- Diverse Image Inpainting with Normalizing Flow (ECCV 2022 · similarity 0.89)
- Image Inpainting via Conditional Texture and Structure Dual Generation (ICCV 2021 · similarity 0.89)
- Contextual-based Image Inpainting: Infer, Match, and Translate (ECCV 2018 · similarity 0.88)
- From Inpainting to Layer Decomposition: Repurposing Generative Inpainting Models for Image Layer Decomposition (CVPR 2026 · similarity 0.87)
- Towards Enhanced Image Inpainting: Mitigating Unwanted Object Insertion and Preserving Color Consistency (CVPR 2025 · similarity 0.86)
- SyFormer: Structure-Guided Synergism Transformer for Large-Portion Image Inpainting (AAAI 2024 · similarity 0.86)
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.
