Research map / 3D vision / Depth and stereo
MER-Tracker: Towards High-Speed 3D Point Tracking via Multi-View Event-RGB Hybrid Cameras
Yiqian Chang, Qinghong Ye, Haoran Xu, Jianing Li, Dongyang Ma, Xuan Wang, Wei Zhang, Yonghong Tian, Peixi Peng
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 event-camera 3D features by direct merging to RGB-only MVTracker (ablation) | MER-Tracker · FMV-Kubric | AJ: 61.9 -> 65.8 | Improved |
| add linear interpolation of RGB 3D features (ablation) | MER-Tracker · FMV-Kubric | AJ: 65.8 -> 68.3 | Improved |
| add anchor (farthest point) sampling after merge (ablation) | MER-Tracker · FMV-Kubric | AJ: 68.3 -> 71.2 | Improved |
| add temporal correlation Transformer (ablation) | MER-Tracker · FMV-Kubric | AJ: 71.2 -> 72.3 | Improved |
| add 1 event camera to 4 RGB cameras (ablation) | MER-Tracker · FMV-Kubric | AJ: 61.9 -> 68.1 | Improved |
| add a second event camera to 4 RGB cameras (ablation) | MER-Tracker · FMV-Kubric | AJ: 68.1 -> 72.3 | Improved |
| use 4 instead of 3 RGB cameras with 2 event cameras (ablation) | MER-Tracker · FMV-Kubric | AJ: 70.6 -> 72.3 | Improved |
| convert event streams to images with E2VID and add as extra views to MVTracker | MVTracker + Frame.Inter · FMV-Kubric | AJ: 62.3 -> 63.5 | Improved |
| frame interpolation (RIFE) instead of frame repetition for MVTracker inputs | MVTracker · FMV-Kubric | AJ: 60.5 -> 62.3 | Improved |
| use MER-Tracker with native event features instead of E2V image conversion | MVTracker-based tracker · Real Object (5 scenes) | RMSE: 0.307 -> 0.228 | Improved |
Closest papers by meaning
- 3D Feature Tracking via Event Camera (CVPR 2024 · similarity 0.87)
- TrackingWorld: World-centric Monocular 3D Tracking of Almost All Pixels (NeurIPS 2025 · similarity 0.86)
- TAPIP3D: Tracking Any Point in Persistent 3D Geometry (NeurIPS 2025 · similarity 0.86)
- DELTA: DENSE EFFICIENT LONG-RANGE 3D TRACKING FOR ANY VIDEO (ICLR 2025 · similarity 0.85)
- TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Tracking (NeurIPS 2026 · similarity 0.85)
- TAPVid-3D: A Benchmark for Tracking Any Point in 3D (NeurIPS 2024 · 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.
