Research map / 3D vision / Depth and stereo
MBA-VO: Motion Blur Aware Visual Odometry
Peidong Liu, Xingxing Zuo, Viktor Larsson, Marc Pollefeys
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 |
|---|---|---|---|
| enable motion blur aware tracking (model exposure-time trajectory) instead of treating blurred images as sharp (ablation) | MBA-VO · ArchVizInterior ArchViz-1 | ATE (m): 0.22 -> 0.0256 | Improved |
| enable motion blur aware tracking instead of treating blurred images as sharp (ablation) | MBA-VO · ArchVizInterior ArchViz-2 | ATE (m): 0.1558 -> 0.0184 | Improved |
| enable motion blur aware tracking instead of treating blurred images as sharp (ablation) | MBA-VO · ArchVizInterior ArchViz-3 | ATE (m): 0.2113 -> 0.0202 | Improved |
| pre-deblur every frame with DeblurGANv2-mobile before running DSO | DSO · ArchVizInterior ArchViz-1 | ATE (m): 0.2132 -> 0.2065 | No clear effect |
| pre-deblur every frame with DeblurGANv2-mobile before running DSO | DSO · ArchVizInterior ArchViz-3 | ATE (m): 0.1286 -> 0.0481 | Improved |
| pre-deblur every frame with DeblurGANv2-mobile before running ORB-SLAM | ORB-SLAM · ArchVizInterior ArchViz-3 | ATE (m): 0.1008 -> 0.0197 | Improved |
| pre-deblur frames with DeblurGANv2-mobile before running ORB-SLAM | ORB-SLAM · ArchVizInterior ArchViz-2 | frame drops (%): 1.068 -> 2.763 | Hurt |
| use SRNDeblurNet instead of DeblurGANv2-mobile for deblurring (ablation) | deblurring network · ArchVizInterior | PSNR (dB): 28.66 -> 30.01 | Improved |
| apply blur aware MBA-VO on top of DSO for blurred images | DSO · ArchVizInterior ArchViz-1 | ATE (m): 0.2132 -> 0.0256 | Improved |
| use MBA-VO instead of DSO on real blurred TUM RGB-D | DSO · TUM RGB-D fr1-desk | ATE (m): 0.4956 -> 0.1021 | Improved |
| use MBA-VO instead of DSO on real blurred TUM RGB-D | DSO · TUM RGB-D fr1-desk2 | ATE (m): 0.7762 -> 0.3997 | Improved |
| use MBA-VO instead of DSO on real blurred TUM RGB-D | DSO · TUM RGB-D fr1-room | ATE (m): 0.2992 -> 0.1435 | Improved |
Closest papers by meaning
- LEAP-VO: Long-term Effective Any Point Tracking for Visual Odometry (CVPR 2024 · similarity 0.89)
- Direct Sparse Odometry (TPAMI 2017 · similarity 0.89)
- VOLDOR: Visual Odometry From Log-Logistic Dense Optical Flow Residuals (CVPR 2020 · similarity 0.89)
- Direct Sparse Odometry With Rolling Shutter (ECCV 2018 · similarity 0.88)
- Leveraging Consistent Spatio-Temporal Correspondence for Robust Visual Odometry (AAAI 2025 · similarity 0.85)
- ZeroVO: Visual Odometry with Minimal Assumptions (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.
