Research map / Image generation / Editing and style transfer
Dual-Path Image Inpainting With Auxiliary GAN Inversion
Check 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 the inversion path to the feed-forward inpainting path (ablation) | Dual-path inpainting network · FFHQ test, Ave% | ℓ1 (%): 2.85 -> 2.17 | Helped |
| Add the inversion path to the feed-forward inpainting path (ablation) | Dual-path inpainting network · FFHQ test, Ave% | SSIM: 0.814 -> 0.882 | Helped |
| Add the inversion path to the feed-forward inpainting path (ablation) | Dual-path inpainting network · FFHQ test, Ave% | FID: 7.85 -> 4.02 | Helped |
| Add the feed-forward path alongside the inversion path (ablation) | Dual-path inpainting network · FFHQ test, Ave% | ℓ1 (%): 5.72 -> 2.17 | Helped |
| Add the feed-forward path alongside the inversion path (ablation) | Dual-path inpainting network · FFHQ test, Ave% | SSIM: 0.773 -> 0.882 | Helped |
| Add deformable feature fusion instead of direct concatenation (ablation) | Dual-path inpainting network · FFHQ test, Ave% | FID: 4.73 -> 4.02 | Helped |
| Use optimization-based rather than learning-based GAN inversion (ablation) | Dual-path inpainting network · FFHQ test | ℓ1 (%): 2.17 -> 2.02 | Helped |
| Use optimization-based rather than learning-based GAN inversion (ablation) | Dual-path inpainting network · FFHQ test | Speed (s/frame): 0.093 -> 45 | Hurt |
| Use hybrid learning- and optimization-based rather than learning-based GAN inversion (ablation) | Dual-path inpainting network · FFHQ test | SSIM: 0.882 -> 0.889 | Helped |
| Use hybrid learning- and optimization-based rather than learning-based GAN inversion (ablation) | Dual-path inpainting network · FFHQ test | Speed (s/frame): 0.093 -> 45 | Hurt |
Closest papers by meaning
- WarpGAN: Warping-Guided 3D GAN Inversion with Style-Based Novel View Inpainting (NeurIPS 2025 · similarity 0.90)
- InOut: Diverse Image Outpainting via GAN Inversion (CVPR 2022 · similarity 0.88)
- Prior Guided GAN Based Semantic Inpainting (CVPR 2020 · similarity 0.88)
- Gradual Residuals Alignment: A Dual-Stream Framework for GAN Inversion and Image Attribute Editing (AAAI 2024 · similarity 0.85)
- In-N-Out: Faithful 3D GAN Inversion with Volumetric Decomposition for Face Editing (CVPR 2024 · similarity 0.85)
- Image Inpainting via Iteratively Decoupled Probabilistic Modeling (ICLR 2024 · 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.