Research map / Probabilistic and scientific modeling / Variational and generative sampling
VOLDOR: Visual Odometry From Log-Logistic Dense Optical Flow Residuals
Zhixiang Min, Yiding Yang, Enrique Dunn
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 |
|---|---|---|---|
| Use VOLDOR instead of MLM-SFM | Monocular VO · KITTI odometry training sequences 0–10 | translation error (%): 2.03 -> 1.32 | Improved |
| Use VOLDOR instead of MLM-SFM | Monocular VO · KITTI odometry training sequences 0–10 | rotation error (deg/m): 0.0045 -> 0.0042 | Improved |
| Use VOLDOR instead of the strongest listed monocular translation baseline | Monocular VO · KITTI odometry test sequences 11–21 | translation error (%): 1.19 -> 1.65 | Hurt |
| Use VOLDOR instead of BVO | Monocular VO · KITTI odometry test sequences 11–21 | rotation error (deg/m): 0.0036 -> 0.0050 | Hurt |
| Use VOLDOR instead of the strongest listed baseline on each sequence | Monocular VO · TUM RGB-D | translation RMSE (m) | Mixed |
| Use Fisk residuals instead of other tested analytic residual models (ablation) | VOLDOR · Depth estimation with few reliable observations | depth EPE | Improved |
| Use Fisk residuals instead of other residual models (ablation) | VOLDOR · Depth estimation with many reliable observations | depth EPE | Improved |
| Use MIE instead of MLE for depth inference (ablation) | VOLDOR with Fisk residuals · Depth estimation across the image sequence | depth accuracy | Improved |
| Use MLE instead of MIE for depth inference (ablation) | VOLDOR · Depth estimation with very low rigidness | depth accuracy | Improved |
| Combine PWC-Net flow, Fisk residuals, and MIE (ablation) | VOLDOR · Camera pose estimation | pose accuracy | Improved |
| Filter pixels to W^j>5 instead of including all pixels with W^j>0 (ablation) | VOLDOR depth estimates · KITTI stereo benchmark | disparity EPE (px): 1.2304 -> 0.5616 | Improved |
| Filter pixels to W^j>5 instead of including all pixels with W^j>0 (ablation) | VOLDOR depth estimates · KITTI stereo benchmark | density: 100.00% -> 27.07% | Hurt |
Closest papers by meaning
- Direct Sparse Odometry (TPAMI 2017 · similarity 0.89)
- MBA-VO: Motion Blur Aware Visual Odometry (ICCV 2021 · similarity 0.89)
- LEAP-VO: Long-term Effective Any Point Tracking for Visual Odometry (CVPR 2024 · similarity 0.88)
- Leveraging Consistent Spatio-Temporal Correspondence for Robust Visual Odometry (AAAI 2025 · similarity 0.86)
- ZeroVO: Visual Odometry with Minimal Assumptions (CVPR 2025 · similarity 0.86)
- UFM: A Simple Path towards Unified Dense Correspondence with Flow (NeurIPS 2025 · 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.
