Research map / Image generation / Editing and style transfer
Fader Networks:Manipulating Images by Sliding Attributes
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 |
|---|---|---|---|
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Mouth | Naturalness: 11.4 -> 79.0 | Helped |
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Smile | Naturalness: 22.9 -> 31.4 | Helped |
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Glasses | Naturalness: 9.6 -> 45.3 | Helped |
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Mouth | Accuracy: 10.1 -> 66.2 | Helped |
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Smile | Accuracy: 9.9 -> 97.1 | Helped |
| Use Fader Networks instead of IcGAN for attribute-swapped images | CelebA attribute editing · CelebA test, Glasses | Accuracy: 47.5 -> 76.6 | Helped |
| Use Fader Networks rather than IcGAN for reconstruction | CelebA attribute editing · CelebA test | RMSE: 0.028 -> 0.0009 | Helped |
| Add discriminator dropout (ablation) | Fader Networks · CelebA | Model performance | Helped |
| Add dropout in the first decoder layers (ablation) | Fader Networks · CelebA | Model performance | Hurt |
| Use a gradual schedule for the discriminator loss weight (ablation) | Fader Networks · CelebA | Model performance | Helped |
| Use eight encoder layers instead of seven (ablation) | Fader Networks · Image attribute editing | Performance | Hurt |
| Use six encoder layers instead of seven (ablation) | Fader Networks · Image attribute editing | Performance | No clear effect |
Closest papers by meaning
- Learning Residual Images for Face Attribute Manipulation (CVPR 2017 · similarity 0.88)
- All-in-One Slider for Attribute Manipulation in Diffusion Models (CVPR 2026 · similarity 0.86)
- Discovering Interpretable Latent Space Directions of GANs Beyond Binary Attributes (CVPR 2021 · similarity 0.85)
- STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing (CVPR 2019 · similarity 0.85)
- Controllable Person Image Synthesis With Attribute-Decomposed GAN (CVPR 2020 · similarity 0.85)
- Multi-Attribute Transfer via Disentangled Representation (AAAI 2019 · 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.