Research map / Point clouds / Segmentation and detection
MLVSNet: Multi-Level Voting Siamese Network for 3D Visual Tracking
Zhoutao Wang, Qian Xie, Yu-Kun Lai, Jing Wu, Kun Long, Jun 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 |
|---|---|---|---|
| replace TSFA with Target-Guided Attention (ablation) | P2B · KITTI 3D tracking (Mean) | Precision: 60.0 -> 62.3 | Improved |
| replace TSFA with Target-Guided Attention (ablation) | P2B · KITTI 3D tracking (Mean) | Success: 42.4 -> 42.3 | No clear effect |
| replace TSFA with TGA for efficiency (ablation) | P2B · KITTI Car | testing speed (fps): 84 -> 93 | Improved |
| replace TSFA with TGA for efficiency (ablation) | P2B · KITTI Car | training time per epoch (min): 21 min -> 16 min | Improved |
| add multi-level voting from last two layers (ablation) | TGA model · KITTI 3D tracking (Mean) | Success: 42.3 -> 44.4 | Improved |
| add multi-level voting from last two layers (ablation) | TGA model · KITTI 3D tracking (Mean) | Precision: 62.3 -> 64.9 | Improved |
| also vote from the shallowest layer (3+2+1 vs 3+2) (ablation) | MLV · KITTI 3D tracking (Mean) | Success: 44.4 -> 37.2 | Hurt |
| also vote from the shallowest layer (3+2+1 vs 3+2) (ablation) | MLV · KITTI 3D tracking (Mean) | Precision: 64.9 -> 55.6 | Hurt |
| vote from layers 3+2 instead of layer 3 only (ablation) | MLV · KITTI 3D tracking (Mean) | Success: 42.3 -> 44.4 | Improved |
| add Vote-cluster Feature Enhancement (ablation) | MLVSNet · KITTI 3D tracking (Mean) | Success: 44.4 -> 45.7 | Improved |
| add Vote-cluster Feature Enhancement (ablation) | MLVSNet · KITTI 3D tracking (Cyclist) | Success: 29.6 -> 34.3 | Improved |
| add Vote-cluster Feature Enhancement (ablation) | MLVSNet · KITTI 3D tracking (Mean) | Precision: 64.9 -> 66.6 | Improved |
Later papers that compared with this method
2 later accepted papers report MLVSNet as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- Exploring Point-BEV Fusion for 3D Point Cloud Object Tracking With Transformer (TPAMI 2024 · better on 1 of 1 shared result)
- MixCycle: Mixup Assisted Semi-Supervised 3D Single Object Tracking with Cycle Consistency (ICCV 2023 · better on 2 of 2 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
- VENet: Voting Enhancement Network for 3D Object Detection (ICCV 2021 · similarity 0.89)
- Deep Hough Voting for 3D Object Detection in Point Clouds (ICCV 2019 · similarity 0.89)
- Back-Tracing Representative Points for Voting-Based 3D Object Detection in Point Clouds (CVPR 2021 · similarity 0.88)
- MLCVNet: Multi-Level Context VoteNet for 3D Object Detection (CVPR 2020 · similarity 0.88)
- 3D Siamese Transformer Network for Single Object Tracking on Point Clouds (ECCV 2022 · similarity 0.88)
- High Performance Visual Tracking With Siamese Region Proposal Network (CVPR 2018 · similarity 0.88)
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.
