Research map / Detection and segmentation / Open-set and few-shot detection
Image Inpainting with Cascaded Modulation GAN and Object-Aware Training
Haitian Zheng, Zhe Lin, Jingwan Lu, Scott Cohen, Eli Shechtman, Connelly Barnes, Jianming Zhang, Ning Xu, Sohrab Amirghodsi, Jiebo Luo
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 masked R1 regularization (ablation) | Baseline · Places2 val | FID: 2.530 -> 2.475 | Improved |
| use spatial modulation instead of global modulation (ablation) | Baseline + mR1 · Places2 val | FID: 2.475 -> 2.398 | Improved |
| cascade global and spatial modulation (ablation) | Baseline + mR1 · Places2 val | FID: 2.475 -> 2.187 | Improved |
| replace second-stage spatial modulation with global modulation (g-g) (ablation) | CM (g-s) · Places2 val | FID: 2.187 -> 2.247 | Hurt |
| replace second-stage modulation with a prior spatial modulation (ablation) | CM (g-s) · Places2 val | FID: 2.187 -> 2.915 | Hurt |
| add spatially-aware demodulation (ablation) | CM (g-s) plain · Places2 val | FID: 2.392 -> 2.187 | Improved |
| add segmentation perceptual loss (ablation) | Baseline + CM · Places2 val | FID: 2.187 -> 1.730 | Improved |
| add segmentation perceptual loss (ablation) | Baseline + CM · Places2 val | P-IDS: 23.86 -> 19.73 | Hurt |
| add Fourier convolution encoder (ablation) | Baseline + CM + perc. · Places2 val | FID: 1.730 -> 1.628 | Improved |
| add object-aware training (ablation) | CoModGAN · Places2 val | FID: 3.724 -> 2.599 | Improved |
| add object-aware training (ablation) | LaMa · Places2 val | FID: 3.864 -> 2.884 | Improved |
| replace CoModGAN with CM-GAN, both with object-aware training (ablation) | CoModGAN + OT · Places2 val | FID: 2.599 -> 1.628 | Improved |
Closest papers by meaning
- High-Fidelity Image Inpainting with GAN Inversion (ECCV 2022 · similarity 0.89)
- Prior Guided GAN Based Semantic Inpainting (CVPR 2020 · similarity 0.89)
- High-Fidelity Image Inpainting with Multimodal Guided GAN Inversion (IJCV 2025 · similarity 0.89)
- Dual-Path Image Inpainting With Auxiliary GAN Inversion (CVPR 2022 · similarity 0.88)
- E$^2$GAN: Efficient Training of Efficient GANs for Image-to-Image Translation (ICML 2024 · similarity 0.87)
- Image Inpainting via Iteratively Decoupled Probabilistic Modeling (ICLR 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.
