Research map / 3D vision / 3D generation
Textureless Deformable Object Tracking With Invisible Markers
Xinyuan Li, Yu Guo, Yubei Tu, Yu Ji, Yanchen Liu, Jinwei Ye, Changxi 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 linear interpolation of 3D features between consecutive UV frames (ablation) | marker-to-reference feature transfer · 4 ms trigger delay grid target | point-to-ray misalignment (mm): 2.23 -> 0.28 | Improved |
| use fluorescent invisible markers for tracking instead of marker-free video | R3DS Wrap4D tracking · paper scene P9 | feature tracking robustness: lost most points within 20 frames of 1800 -> reliable over time | Improved |
| fine-tune pre-trained RAFT on DOT paper scenes | RAFT (Flying Chairs pretrained) · P2 and DeSurT | flow error | Improved |
| feed DOT fine-tuned flow correspondences into RoBuSfT | RoBuSfT template estimation · textureless paper | template reconstruction | Improved |
| use PWC-Net instead of GF on textured paper | optical flow · DOT paper scenes | flow error (pixels): 4.662 -> 0.441 | Improved |
| use learned flow DALF instead of GF on cloth | optical flow · DOT cloth scenes | flow error (pixels): 2.221 -> 5.835 | Hurt |
| use RAFT instead of PWC-Net on white paper P1 | optical flow · DOT paper P1 | flow error (pixels): 8.215 -> 196.2 | Hurt |
| textureless cloth and white paper | all benchmarked flow methods · DOT cloth and P1 | flow error | Hurt |
| use InterWild instead of InterNet for hand reconstruction | template hand reconstruction · DOT hand scenes | RMSE on warped image: 41.92 -> 17.75 | Improved |
| use different dye colors for hand and object | template fitting in hand-object interaction · DOT hand scenes | fitting accuracy | Improved |
| tune dye concentration to 1/625 (ablation) | fluorescent markers · marker detection | detection robustness | Mixed |
Closest papers by meaning
- Deformable Surface Tracking by Graph Matching (ICCV 2019 · similarity 0.88)
- Tracking by 3D Model Estimation of Unknown Objects in Videos (ICCV 2023 · similarity 0.87)
- 4DComplete: Non-Rigid Motion Estimation Beyond the Observable Surface (ICCV 2021 · similarity 0.86)
- MoDA: Modeling Deformable 3D Objects from Casual Videos (IJCV 2024 · similarity 0.85)
- Dense Optical Tracking: Connecting the Dots (CVPR 2024 · similarity 0.84)
- Video Diffusion Models Excel at Tracking Similar-Looking Objects Without Supervision (NeurIPS 2025 · similarity 0.84)
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.
