Research map / Image generation / Super-resolution
Super-Resolving Cross-Domain Face Miniatures by Peeking at One-Shot Exemplar
Peike Li, Xin Yu, Yi Yang
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 multi-scale adaptive latent encoding features (ablation) | DAP-FSR baseline network · CelebA->ExtendedYaleB | FIQ: 0.28 -> 0.30 | Improved |
| add multi-scale features (ablation) | DAP-FSR baseline network · CelebA->ExtendedYaleB | SSIM: 0.44 -> 0.44 | No clear effect |
| add multiple ISTN alignment modules (ablation) | DAP-FSR with multi-scale features · CelebA->ExtendedYaleB | PSNR: 17.82 -> 18.41 | Improved |
| predict offset scale w.r.t. latent mean and variance (manifold constraint) (ablation) | DAP-FSR with ISTN · CelebA->ExtendedYaleB | PSNR: 18.41 -> 20.39 | Improved |
| predict offset scale w.r.t. latent mean and variance (ablation) | DAP-FSR with ISTN · CelebA->ExtendedYaleB | FIQ: 0.34 -> 0.41 | Improved |
| add style-mixing augmentation from one-shot exemplar (ablation) | DAP-FSR source-only · CelebA->ExtendedYaleB | PSNR: 20.39 -> 21.97 | Improved |
| add style-mixing augmentation from one-shot exemplar (ablation) | DAP-FSR source-only · CelebA->ExtendedYaleB | FIQ: 0.41 -> 0.41 | No clear effect |
| learn soft layer-wise mixing weight instead of replacing last three layers (ablation) | DAP-FSR with style mixing · CelebA->ExtendedYaleB | PSNR: 21.97 -> 22.10 | Improved |
| constrain fine-tuning to AdaIN affine parameters only (ablation) | DAP-FSR with soft mixing · CelebA->ExtendedYaleB | FIQ: 0.42 -> 0.46 | Improved |
| use proposed latent mixing instead of ASM augmentation (ablation) | DAP-FSR adaptation · CelebA->ExtendedYaleB | PSNR: 20.71 -> 22.32 | Improved |
| use proposed latent mixing instead of direct fine-tuning (ablation) | DAP-FSR adaptation · CelebA->ExtendedYaleB | FIQ: 0.30 -> 0.46 | Improved |
| apply ASM style-transfer augmentation and fine-tune baseline FSR model | MTDN · CelebA->ExtendedYaleB | FIQ: 0.32 -> 0.27 | Hurt |
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.
- PULSE — PULSE: Self-Supervised Photo Upsampling via Latent Space Exploration of Generative Models (CVPR 2020) · 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
- FSRNet: End-to-End Learning Face Super-Resolution With Facial Priors (CVPR 2018 · similarity 0.90)
- Face Super-resolution Guided by Facial Component Heatmaps (ECCV 2018 · similarity 0.88)
- Extreme Face Super-Resolution through Identity Fitting and Decoupling (ECCV 2026 · similarity 0.88)
- Super-Resolving Very Low-Resolution Face Images With Supplementary Attributes (CVPR 2018 · similarity 0.88)
- Low-Light Face Super-resolution via Illumination, Structure, and Texture Associated Representation (AAAI 2024 · similarity 0.87)
- IODA: Instance-Guided One-shot Domain Adaptation for Super-Resolution (NeurIPS 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.
