Research map / 3D vision / Human pose estimation
SRT3D: A Sparse Region-Based 3D Object Tracking Approach for the Real World
Manuel Stoiber, Martin Pfanne, Klaus H. Strobl, Rudolph Triebel, Alin Albu‐Schäffer
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 |
|---|---|---|---|
| upgrade from previous sparse approach to SRT3D (extended smoothed step function, new optimization) | sparse region-based tracker · RBOT (regular) | tracking success rate (avg): 90.0 -> 94.2 | Improved |
| upgrade from previous sparse approach to SRT3D | sparse region-based tracker · RBOT (noise) | tracking success rate (avg): 71.5 -> 81.7 | Improved |
| upgrade from previous sparse approach to SRT3D | sparse region-based tracker · RBOT (unmodeled occlusion) | tracking success rate (avg): 85.6 -> 93.2 | Improved |
| upgrade from previous sparse approach to SRT3D | sparse region-based tracker · RBOT (modeled occlusion) | tracking success rate (avg): 89.1 -> 94.1 | Improved |
| add explicit modeling of known occlusions (occlusion mask rendering) (ablation) | SRT3D · RBOT (occlusion sequence) | tracking success rate (avg): 93.2 -> 94.1 | Improved |
| add occlusion modeling (render occlusion masks, track two objects) (ablation) | SRT3D · RBOT | runtime per frame (ms): 1.1 -> 5.1 | Hurt |
| upgrade from previous approach to SRT3D under occlusion modeling | sparse region-based tracker · RBOT (modeled occlusion) | runtime per frame (ms): 7.4 -> 5.1 | Improved |
| upgrade from previous approach to SRT3D without occlusion modeling | sparse region-based tracker · RBOT (no occlusion modeling) | runtime per frame (ms): 1.0 -> 1.1 | No clear effect |
| use sparse correspondence-line SRT3D instead of dense region-based method | region-based tracking · RBOT (regular) | tracking success rate (avg): more than 14 percentage points better than the next best, dense, region-based approach | Improved |
Later papers that compared with this method
3 later accepted papers report SRT3D as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Deep Active Contours for Real-time 6-DoF Object Tracking (ICCV 2023 · better on 2 of 2 shared results)
- Large-Displacement 3D Object Tracking with Hybrid Non-local Optimization (ECCV 2022 · better on 4 of 4 shared results)
- Iterative Corresponding Geometry: Fusing Region and Depth for Highly Efficient 3D Tracking of Textureless Objects (CVPR 2022 · 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
- SpatialTrackerV2: Advancing 3D Point Tracking with Explicit Camera Motion (ICCV 2025 · similarity 0.90)
- M3SOT: Multi-Frame, Multi-Field, Multi-Space 3D Single Object Tracking (AAAI 2024 · similarity 0.89)
- Prior-free 3D Object Tracking (CVPR 2025 · similarity 0.87)
- TrackCraft3R: Repurposing Video Diffusion Transformers for Dense 3D Tracking (NeurIPS 2026 · similarity 0.87)
- SPAR3D: Stable Point-Aware Reconstruction of 3D Objects from Single Images (CVPR 2025 · similarity 0.87)
- TAPVid-3D: A Benchmark for Tracking Any Point in 3D (NeurIPS 2024 · 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.
