SparseSplat uses entropy-based probabilistic sampling and a specialized point cloud network to generate compact 3D Gaussian maps that retain high rendering quality with far fewer Gaussians than prior feed-forward methods.
3.3, our 3D-Local Attribute Prediction framework employs a lightweight predictor to regress Gaus- sian attributes based on K-nearest neighbors in 3D space
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
-
SparseSplat: Towards Applicable Feed-Forward 3D Gaussian Splatting with Pixel-Unaligned Prediction
SparseSplat uses entropy-based probabilistic sampling and a specialized point cloud network to generate compact 3D Gaussian maps that retain high rendering quality with far fewer Gaussians than prior feed-forward methods.