Research map / Image generation / Super-resolution
FSRNet: End-to-End Learning Face Super-Resolution With Facial Priors
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 facial landmark heatmaps as prior information (ablation) | Face SR baseline · Helen | PSNR: +0.4 dB | Helped |
| Add parsing maps as prior information (ablation) | Face SR baseline · Helen | PSNR: +1.0 dB | Helped |
| Add both landmark heatmaps and parsing maps (ablation) | Face SR baseline · Helen | PSNR: +1.05 dB | Helped |
| Add an unsupervised prior-estimation branch without prior-label supervision (ablation) | FSRNet · Face SR | PSNR | Helped |
| Supervise the prior-estimation branch with facial prior labels (ablation) | FSRNet · Face SR | PSNR | Helped |
| Use both landmark and parsing priors instead of either alone (ablation) | FSRNet · Face SR | PSNR | Helped |
| Increase the prior-estimation network from 1 to 2 hourglasses (ablation) | FSRNet · CelebA | PSNR: 25.69 -> 25.87 dB | Helped |
| Increase the prior-estimation network from 2 to 4 hourglasses (ablation) | FSRNet · CelebA | PSNR: 25.87 -> 25.95 dB | Helped |
| Fuse multiple augmented test-image outputs at inference | FSRNet · Helen | PSNR: 25.87 -> 26.21 | Helped |
| Add adversarial and perceptual losses | FSRNet · Helen | PSNR: 25.87 -> 25.10 | Hurt |
| Add adversarial and perceptual losses | FSRNet · Helen | NRMSE: 4.18 -> 3.97 | Helped |
| Use global rather than local parsing maps (ablation) | Face SR with parsing priors · Helen | PSNR | Helped |
Closest papers by meaning
- Deep Face Super-Resolution With Iterative Collaboration Between Attentive Recovery and Landmark Estimation (CVPR 2020 · similarity 0.92)
- Super-FAN: Integrated Facial Landmark Localization and Super-Resolution of Real-World Low Resolution Faces in Arbitrary Poses With GANs (CVPR 2018 · similarity 0.91)
- Spatial-Frequency Mutual Learning for Face Super-Resolution (CVPR 2023 · similarity 0.89)
- Joint Super-Resolution and Alignment of Tiny Faces (AAAI 2020 · similarity 0.89)
- GCFSR: A Generative and Controllable Face Super Resolution Method Without Facial and GAN Priors (CVPR 2022 · similarity 0.89)
- Low-Light Face Super-resolution via Illumination, Structure, and Texture Associated Representation (AAAI 2024 · similarity 0.87)
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.