Research map / 3D vision / 3D generation
Towards Generic 3D Tracking in RGBD Videos: Benchmark and Baseline
Jinyu Yang, Zhongqun Zhang, Zhe Li, Hyung Jin Chang, Aleš Leonardis, Feng Zheng
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 RGBD fusion head to P2B (ablation) | P2B ft · Track-it-in-3D test | Success: 9.4% -> 24.5% | Improved |
| add RGBD fusion head to P2B (ablation) | P2B ft · Track-it-in-3D test | Precision: 8.4% -> 28.2% | Improved |
| add RGBD fusion head to BAT (ablation) | BAT ft · Track-it-in-3D test | Success: 2.5% -> 18.1% | Improved |
| add RGBD fusion head to BAT (ablation) | BAT ft · Track-it-in-3D test | Precision: 4.7% -> 26.0% | Improved |
| use proposed TrackIt3D design (RGBD fusion plus 3D xcorr) over best RGBD-fused voting head (ablation) | P2B++ · Track-it-in-3D test | Success: 24.5% -> 31.1% | Improved |
| add template feature to cross-correlation (ablation) | TrackIt3D · Track-it-in-3D test | Success: 7.0% -> 31.1% | Improved |
| add template feature to cross-correlation (ablation) | TrackIt3D · Track-it-in-3D test | Precision: 5.0% -> 35.0% | Improved |
| add similarity map to 3D cross-correlation (ablation) | TrackIt3D · Track-it-in-3D test | Precision: 33.1% -> 35.0% | Improved |
| add similarity map to 3D cross-correlation (ablation) | TrackIt3D · Track-it-in-3D test | Success: 30.9% -> 31.1% | No clear effect |
| use 3D cross-correlation instead of 2D correlation (ablation) | TrackIt3D · Track-it-in-3D test | Success / Precision: Success 28.3% -> 31.1%, Precision 38.4% -> 35.0% | Mixed |
| finetune P2B on Track-it-in-3D training set | P2B · Track-it-in-3D test | Success: 4.2% -> 9.4% | Improved |
| finetune BAT on Track-it-in-3D training set | BAT · Track-it-in-3D test | Success: 2.5% -> 2.5% | No clear effect |
Closest papers by meaning
- Seeing Behind Objects for 3D Multi-Object Tracking in RGB-D Sequences (CVPR 2021 · similarity 0.91)
- TAPIP3D: Tracking Any Point in Persistent 3D Geometry (NeurIPS 2025 · similarity 0.88)
- TrackingWorld: World-centric Monocular 3D Tracking of Almost All Pixels (NeurIPS 2025 · similarity 0.88)
- TAPVid-3D: A Benchmark for Tracking Any Point in 3D (NeurIPS 2024 · similarity 0.87)
- TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Tracking (NeurIPS 2026 · similarity 0.87)
- SpatialTracker: Tracking Any 2D Pixels in 3D Space (CVPR 2024 · similarity 0.87)
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.
