A Gaussian splatting CBCT method with a compact quadratic Bezier material model reduces metal artifacts about 3 to 5 times faster than neural field baselines while better preserving fine structure.
Unsupervised Polychromatic Neural Representation for
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
-
Splat-Based Metal Artifact Reduction in Cone-Beam CT via Compact Attenuation Modeling
A Gaussian splatting CBCT method with a compact quadratic Bezier material model reduces metal artifacts about 3 to 5 times faster than neural field baselines while better preserving fine structure.