Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2512.22295 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909976694358016 |
|---|---|
| author | Guo, Tian Yuan, Hui Xu, Philip Elizondo, David |
| author_facet | Guo, Tian Yuan, Hui Xu, Philip Elizondo, David |
| contents | We propose SirenPose, a novel loss function that combines the periodic activation properties of sinusoidal representation networks with geometric priors derived from keypoint structures to improve the accuracy of dynamic 3D scene reconstruction. Existing approaches often struggle to maintain motion modeling accuracy and spatiotemporal consistency in fast moving and multi target scenes. By introducing physics inspired constraint mechanisms, SirenPose enforces coherent keypoint predictions across both spatial and temporal dimensions. We further expand the training dataset to 600,000 annotated instances to support robust learning. Experimental results demonstrate that models trained with SirenPose achieve significant improvements in spatiotemporal consistency metrics compared to prior methods, showing superior performance in handling rapid motion and complex scene changes. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_22295 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Learning Dynamic Scene Reconstruction with Sinusoidal Geometric Priors Guo, Tian Yuan, Hui Xu, Philip Elizondo, David Computer Vision and Pattern Recognition We propose SirenPose, a novel loss function that combines the periodic activation properties of sinusoidal representation networks with geometric priors derived from keypoint structures to improve the accuracy of dynamic 3D scene reconstruction. Existing approaches often struggle to maintain motion modeling accuracy and spatiotemporal consistency in fast moving and multi target scenes. By introducing physics inspired constraint mechanisms, SirenPose enforces coherent keypoint predictions across both spatial and temporal dimensions. We further expand the training dataset to 600,000 annotated instances to support robust learning. Experimental results demonstrate that models trained with SirenPose achieve significant improvements in spatiotemporal consistency metrics compared to prior methods, showing superior performance in handling rapid motion and complex scene changes. |
| title | Learning Dynamic Scene Reconstruction with Sinusoidal Geometric Priors |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.22295 |