Research map / Image generation / Image restoration
High-Fidelity Pluralistic Image Completion With Transformers
Ziyu Wan, Jingbo Zhang, Dongdong Chen, Jing Liao
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 transformer appearance prior plus guided CNN upsampling instead of PIC's fully convolutional variational model | image completion · ImageNet, mask 40%-60% | FID: 101.278 -> 59.486 | Improved |
| use transformer appearance prior plus guided CNN upsampling instead of PIC's fully convolutional variational model | image completion · ImageNet, mask 20%-40% | FID: 47.750 -> 28.818 | Improved |
| use transformer appearance prior plus guided CNN upsampling instead of PIC's fully convolutional variational model | image completion · ImageNet, random masks | FID: 59.428 -> 35.842 | Improved |
| use proposed method instead of best baseline EC | image completion · FFHQ, mask 20%-40% | FID: 14.276 -> 10.995 | Improved |
| switch from top-50 sampling to deterministic top-1 sampling (ablation) | Ours · FFHQ, mask 20%-40% | FID: 10.995 -> 10.515 | Improved |
| switch from top-50 sampling to deterministic top-1 sampling (ablation) | Ours · FFHQ, mask 40%-60% | FID: 20.024 -> 20.843 | Hurt |
| switch from top-50 sampling to deterministic top-1 sampling (ablation) | Ours · Places2, random masks | FID: 25.420 -> 25.985 | Hurt |
| switch from top-50 sampling to deterministic top-1 sampling (ablation) | Ours · Places2, mask 40%-60% | PSNR: 22.215 -> 22.635 | Improved |
| remove instance normalization from the upsampling network (ablation) | guided upsampling CNN · image completion | visual quality | Improved |
| add transformer prior versus fully convolutional methods on geometric primitive completion | image completion · pentagram shape completion (ImageNet-trained) | qualitative reconstruction | Improved |
| use proposed method versus other baselines in a user study | image completion · user study ranking, 30 images, 28 participants | top-1 rank ratio: 73.70% | Improved |
Closest papers by meaning
- High-Fidelity and Efficient Pluralistic Image Completion With Transformers (TPAMI 2024 · similarity 0.99)
- Don't Look into the Dark: Latent Codes for Pluralistic Image Inpainting (CVPR 2024 · similarity 0.85)
- Large Occluded Human Image Completion via Image-Prior Cooperating (AAAI 2024 · similarity 0.85)
- MaGIC: Multi-modality Guided Image Completion (ICLR 2024 · similarity 0.84)
- CORP: Closed-Form One-shot Representation-Preserving Structured Pruning for Transformers (NeurIPS 2026 · similarity 0.84)
- Completion as Enhancement: A Degradation-Aware Selective Image Guided Network for Depth Completion (CVPR 2025 · similarity 0.84)
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.
