Research map / Image generation / Super-resolution
Texture Hallucination for Large-Factor Painting Super-Resolution
Yulun Zhang, Zhifei Zhang, Stephen DiVerdi, Zhaowen Wang, Jose Echevarria, Yun Fu
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 wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · PaintHD 8x | PI: 5.253 -> 3.670 | Improved |
| add wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · PaintHD 16x | PI: 5.815 -> 4.131 | Improved |
| add wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · CUFED5 8x | PI: 4.520 -> 2.339 | Improved |
| add wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · CUFED5 16x | PI: 5.298 -> 2.499 | Improved |
| add wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · PaintHD 8x | PSNR (dB): 24.02 -> 22.49 | Hurt |
| add wavelet texture, degradation, perceptual and adversarial losses to reconstruction-only training (ablation) | Ours · CUFED5 8x | PSNR (dB): 22.40 -> 20.36 | Hurt |
| use simplified pipeline instead of SRNTT with reconstruction loss only (ablation) | Ours-Lrec · PaintHD 8x | PI: 7.992 -> 5.253 | Improved |
| use simplified pipeline instead of SRNTT with reconstruction loss only (ablation) | Ours-Lrec · PaintHD 16x | PI: 8.464 -> 5.815 | Improved |
| use simplified pipeline instead of SRNTT with reconstruction loss only (ablation) | Ours-Lrec · PaintHD 8x | PSNR (dB): 23.96 -> 24.02 | Improved |
| apply texture loss on wavelet high-frequency band instead of raw features (ablation) | Ours · PaintHD 8x | visual quality (qualitative) | Improved |
| add degradation loss to reconstruction loss (ablation) | Ours · PaintHD 8x | visual quality (qualitative) | Improved |
| use reference-based SR with texture transfer instead of SISR RCAN | Ours · PaintHD 8x | PI: 7.448 -> 3.670 | Improved |
Closest papers by meaning
- Image Super-Resolution by Neural Texture Transfer (CVPR 2019 · similarity 0.87)
- Task Decoupled Framework for Reference-Based Super-Resolution (CVPR 2022 · similarity 0.87)
- Training Generative Image Super-Resolution Models by Wavelet-Domain Losses Enables Better Control of Artifacts (CVPR 2024 · similarity 0.86)
- Learning Texture Transformer Network for Image Super-Resolution (CVPR 2020 · similarity 0.86)
- PBR-SR: Mesh PBR Texture Super Resolution from 2D Image Priors (NeurIPS 2025 · similarity 0.85)
- Neural Super-Resolution for Real-time Rendering with Radiance Demodulation (CVPR 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.
