Research map / Image generation / GANs
MI-GAN: A Simple Baseline for Image Inpainting on Mobile Devices
Andranik Sargsyan, Shant Navasardyan, Xingqian Xu, Humphrey Shi
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 online model re-parametrization to the baseline (ablation) | MI-GAN 256 · Places2 256 | FID: 14.94 -> 12.22 | Improved |
| add knowledge distillation from Co-Mod-GAN on painting-branch outputs (ablation) | MI-GAN 256 with re-param · Places2 256 | FID: 12.22 -> 11.83 | Improved |
| add re-parametrization and distillation to baseline (ablation) | MI-GAN 256 · Places2 256 | LPIPS: 0.407 -> 0.39 | Improved |
| MI-GAN versus LaMa on Places2 256 FID | MI-GAN · Places2 256 | FID: 22.00 -> 11.83 | Improved |
| MI-GAN versus Co-Mod-GAN and SH-GAN on FID | MI-GAN · Places2 256 | FID: 9.32 -> 11.83 | Hurt |
| use MI-GAN instead of Co-Mod-GAN on mobile CPUs | MI-GAN · iPhone7 256 | inference time (ms): 4475.33 -> 1030.25 | Improved |
| MI-GAN versus teacher-class models at 512 resolution | MI-GAN 512 · Places2 512 | FID: 7.03 -> 10.00 | Mixed |
Closest papers by meaning
- Deep Learning-Based Image and Video Inpainting: A Survey (IJCV 2024 · 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)
- The GAN is dead; long live the GAN! A Modern GAN Baseline (NeurIPS 2024 · similarity 0.86)
- InstantPainting: Expanding GANs for Efficient Text-Conditioned Image Generation Platform (AAAI 2025 · similarity 0.86)
- TurboFill: Adapting Few-step Text-to-image Model for Fast Image Inpainting (CVPR 2025 · 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.
