A feed-forward transformer predicts SMPL-X vertex-aligned 3D Gaussians in a canonical T-pose, enabling real-time animation by linear blend skinning without per-frame network inference.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Real-Time Human Reconstruction and Animation using Feed-Forward Gaussian Splatting
A feed-forward transformer predicts SMPL-X vertex-aligned 3D Gaussians in a canonical T-pose, enabling real-time animation by linear blend skinning without per-frame network inference.