Research map / 3D vision / Depth and stereo
SceneTracker: Long-Term Scene Flow Estimation Network
Bo Wang, Jian Li, Yang Yu, Li Liu, Zhenping Sun, Dewen Hu
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 depth residual feature (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 91.43 -> 95.41 | Improved |
| add depth residual feature (ablation) | SceneTracker · LSFOdyssey test | EPE_3D: 0.085 -> 0.058 | Improved |
| update template features during tracking (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 95.27 -> 95.41 | Improved |
| inference window size 16 instead of 8 (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 93.32 -> 95.41 | Improved |
| inference window size 24 instead of 16 (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 95.41 -> 81.47 | Hurt |
| 6x2 Transformer blocks instead of 4x2 (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 94.93 -> 95.41 | Improved |
| 8x2 instead of 6x2 Transformer blocks (ablation) | SceneTracker · LSFOdyssey test | delta_3D^avg: 95.41 -> 94.10 | Hurt |
| add local and global auxiliary points | SceneTracker ('One' mode) · LSFDriving pedestrian | EPE_3D: 2.902 -> 0.707 | Improved |
| add global auxiliary points to 'All' inference | SceneTracker · LSFDriving pedestrian | EPE_3D: 2.890 -> 0.486 | Improved |
| use SceneTracker instead of SF baseline | SceneTracker · LSFOdyssey test | MAE_3D: 0.176 -> 0.052 | Improved |
| use SceneTracker instead of TAP baseline | SceneTracker · LSFOdyssey test | EPE_3D: 0.573 -> 0.058 | Improved |
| SceneTracker vs SpatialTracker | SceneTracker · LSFDriving background | MAE_2D: 1.43 -> 1.85 | Hurt |
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.
- SpatialTracker — SpatialTracker: Tracking Any 2D Pixels in 3D Space (CVPR 2024) · better on 5 of 5 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
- Deep Scene Flow Learning: From 2D Images to 3D Point Clouds (TPAMI 2023 · similarity 0.89)
- DifFlow3D: Hierarchical Diffusion Models for Uncertainty-Aware 3D Scene Flow Estimation (TPAMI 2025 · similarity 0.88)
- MESSENGER: Memory-Enhanced Sequential Scene Flow Estimation via Autoregressive Next-Frame Forecasting (NeurIPS 2026 · similarity 0.88)
- DifFlow3D: Toward Robust Uncertainty-Aware Scene Flow Estimation with Iterative Diffusion-Based Refinement (CVPR 2024 · similarity 0.87)
- DeltaFlow: An Efficient Multi-frame Scene Flow Estimation Method (NeurIPS 2025 · similarity 0.87)
- Zero-Shot Monocular Scene Flow Estimation in the Wild (CVPR 2025 · 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.
