Research map / 3D vision / Face and shape
DeFMO: Deblurring and Shape Recovery of Fast Moving Objects
Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari, Jiri Matas, Marc Pollefeys
Check 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 image reconstruction loss (ablation) | DeFMO · Falling Objects test | PSNR: 24.8 -> 26.4 | Improved |
| Add supervised appearance reconstruction loss (ablation) | DeFMO · Falling Objects test | PSNR: 19.7 -> 26.4 | Improved |
| Add time-consistency loss (ablation) | DeFMO · Falling Objects test | PSNR: 25.8 -> 26.4 | Improved |
| Add latent learning loss (ablation) | DeFMO · Falling Objects test | PSNR: 26.1 -> 26.4 | Improved |
| Add sharpness loss (ablation) | DeFMO · Falling Objects test | TIoU: 0.676 -> 0.703 | Improved |
| Train with only supervised appearance reconstruction loss (ablation) | DeFMO · Falling Objects test | PSNR: 24.8 -> 19.7 | Hurt |
| Remove time-consistency, latent learning, and sharpness losses while retaining appearance and image reconstruction losses (ablation) | DeFMO · Falling Objects test | TIoU: 0.703 -> 0.653 | Hurt |
| Remove sharpness loss (ablation) | DeFMO · Falling Objects test | PSNR: 26.4 -> 26.4 | No clear effect |
| Use DeFMO instead of TbD-3D-Oracle | Falling Objects · Falling Objects | PSNR: 23.38 -> 26.83 | Improved |
| Use DeFMO instead of TbD-3D-Oracle | TbD-3D · TbD-3D | PSNR: 24.84 -> 26.23 | Improved |
| Use DeFMO instead of TbD-3D-Oracle | TbD · TbD | SSIM: 0.712 -> 0.602 | Hurt |
Later papers that compared with this method
1 later accepted paper reports DeFMO as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Shape from Blur: Recovering Textured 3D Shape and Motion of Fast Moving Objects (NeurIPS 2021 · better on 9 of 9 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Shape from Blur: Recovering Textured 3D Shape and Motion of Fast Moving Objects (NeurIPS 2021 · similarity 0.88)
- Motion-From-Blur: 3D Shape and Motion Estimation of Motion-Blurred Objects in Videos (CVPR 2022 · similarity 0.87)
- GS-Blur: A 3D Scene-Based Dataset for Realistic Image Deblurring (NeurIPS 2024 · similarity 0.86)
- Decompositional Neural Scene Reconstruction with Generative Diffusion Prior (CVPR 2025 · similarity 0.86)
- DynaMoDe-NeRF: Motion-aware Deblurring Neural Radiance Field for Dynamic Scenes (CVPR 2025 · similarity 0.86)
- Real-World Efficient Blind Motion Deblurring via Blur Pixel Discretization (CVPR 2024 · similarity 0.86)
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.
