Research map / Point clouds / Registration and matching
DVLO: Deep Visual-LiDAR Odometry with Local-to-Global Feature Fusion and Bi-Directional Structure Alignment
Jiuming Liu, Dong Zhuo, Zhiheng Feng, Siting Zhu, Chensheng Peng, Zhe Liu, Hesheng Wang
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 clustering-based Local Fuser to the Global Fuser only model (ablation) | DVLO · KITTI odometry (mean 07-10) | t_rel: 1.00 -> 0.82 | Improved |
| add Global Fuser to the Local Fuser only model (ablation) | DVLO · KITTI odometry (mean 07-10) | t_rel: 0.93 -> 0.82 t_rel; 0.47 -> 0.41 r_rel | Improved |
| use clustering-based local fusion instead of attention-based (ablation) | DVLO · KITTI odometry (mean 07-10) | t_rel and inference time: 0.96 -> 0.82 t_rel; 183.76 ms -> 98.50 ms | Improved |
| use clustering-based local fusion instead of convolution-based (ablation) | DVLO · KITTI odometry (mean 07-10) | t_rel: 0.99 -> 0.82 t_rel; 87.24 ms -> 98.50 ms | Mixed |
| use DVLO visual-LiDAR fusion instead of EfficientLO (LiDAR only) | KITTI odometry · KITTI odometry (mean 07-10) | r_rel: 0.41 -> 0.41; t_rel 0.86 -> 0.82 | No clear effect |
| use DVLO fusion module for scene flow estimation | CamLiRAFT · FlyingThings3D val | EPE_3D: 0.049 -> 0.048 | Improved |
| use DVLO trained on 00-06 versus H-VLO trained on 00-08 | KITTI 09-10 · KITTI odometry (mean 09-10) | t_rel: 1.64 -> 0.87 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TransLO — TransLO: A Window-Based Masked Point Transformer Framework for Large-Scale LiDAR Odometry (AAAI 2023) · better on 1 of 1 shared result
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
- TransLO: A Window-Based Masked Point Transformer Framework for Large-Scale LiDAR Odometry (AAAI 2023 · similarity 0.88)
- SDV-LOAM: Semi-Direct Visual–LiDAR Odometry and Mapping (TPAMI 2023 · similarity 0.88)
- Efficient 3D Deep LiDAR Odometry (TPAMI 2022 · similarity 0.87)
- StreamVLO: Streaming Visual-LiDAR Odometry with Cumulative Drift Compensation (CVPR 2026 · similarity 0.87)
- GBlobs: Explicit Local Structure via Gaussian Blobs for Improved Cross-Domain LiDAR-based 3D Object Detection (CVPR 2025 · similarity 0.87)
- LightLoc: Learning Outdoor LiDAR Localization at Light Speed (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.
