A new method using transformer-based triangulation, Gröbner basis algebraic loss, and temporal equivariance achieves state-of-the-art results for uncalibrated multi-view 3D human pose estimation and narrows the gap to calibrated systems.
Transfusion: Cross-view fusion with transformer for 3d human pose estimation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Unconstrained Multi-view Human Pose Estimation with Algebraic Priors
A new method using transformer-based triangulation, Gröbner basis algebraic loss, and temporal equivariance achieves state-of-the-art results for uncalibrated multi-view 3D human pose estimation and narrows the gap to calibrated systems.