Research map / Image generation / GANs
Eye In-Painting With Exemplar Generative Adversarial Networks
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 a reference image as an exemplar | Non-exemplar GAN · Internal benchmark | FID: 11.27 -> 7.67 | Helped |
| Add a perceptual eye code as an exemplar | Non-exemplar GAN · Internal benchmark | FID: 11.27 -> 8.49 | Helped |
| Add a reference image as an exemplar | Non-exemplar GAN · Celeb-ID | FID: 15.30 -> 15.66 | Hurt |
| Add a perceptual eye code as an exemplar | Non-exemplar GAN · Celeb-ID | FID: 15.30 -> 14.62 | Helped |
| Add eye codes | Non-exemplar GAN · Eye in-painting training | content loss: decreased by 8% | Helped |
| Add reference images | Non-exemplar GAN · Eye in-painting training | L1 loss: decreased by 17% | Helped |
| Use perceptual loss (ablation) | Exemplar GAN · Eye in-painting | perceptual quality | No clear effect |
| Use the eye code directly in the generator instead of relying on perceptual loss (ablation) | Code-based Exemplar GAN · Eye in-painting | perceptual quality | Helped |
| Train longer, from epoch 10 to epoch 18 (ablation) | Same eye in-painting network · Eye in-painting example | FID: 7.67 -> 10.55 | Hurt |
| Train longer, from epoch 10 to epoch 18 (ablation) | Same eye in-painting network · Eye in-painting example | L1: 0.01457 -> 0.01386 | Helped |
Closest papers by meaning
- APDrawingGAN: Generating Artistic Portrait Drawings From Face Photos With Hierarchical GANs (CVPR 2019 · similarity 0.87)
- Assessing Eye Aesthetics for Automatic Multi-Reference Eye In-Painting (CVPR 2020 · similarity 0.86)
- Navigating the GAN Parameter Space for Semantic Image Editing (CVPR 2021 · similarity 0.86)
- Editing in Style: Uncovering the Local Semantics of GANs (CVPR 2020 · similarity 0.86)
- High-Resolution Image Synthesis and Semantic Manipulation With Conditional GANs (CVPR 2018 · similarity 0.86)
- Image Inpainting via Iteratively Decoupled Probabilistic Modeling (ICLR 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.