From Sparse to Dense: Spatio-Temporal Fusion for Multi-View 3D Human Pose Estimation with DenseWarper Accepted at ICLR 2026
Ling Li, Changjie Chen, Yuyan Wang, Jiaqing Lyu, Kenglun Chang, Yiyun Chen, Zhidong Deng · 2026
Read From Sparse to Dense: Spatio-Temporal Fusion for Multi-View 3D Human Pose Estimation with DenseWarper Accepted at ICLR 2026 by Ling Li, Changjie Chen, Yuyan Wang, Jiaqing Lyu, Kenglun Chang, Yiyun Chen, Zhidong Deng (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-14525.txt.
Description
Li, Chen, Wang, Lyu, Chang, Chen and Deng feed a pose estimator with staggered views from different cameras at different moments, instead of simultaneous frames. Their DenseWarper model swaps heatmaps across space and time using epipolar geometry, the geometry linking views of the same scene. The authors claim sparse interleaved input beats dense multi-view input on two standard benchmarks and lifts the output frame rate with more cameras.
Catalog record
| CREATOR | Ling Li, Changjie Chen, Yuyan Wang, Jiaqing Lyu, Kenglun Chang, Yiyun Chen, Zhidong Deng |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 62725 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.14525