UGD learns a Gaussian Mixture Model from clean point cloud patches to evaluate denoising degradation without requiring ground-truth clean data.
Robust moving least-squares fitting with sharp features,
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
-
UGD: An Unsupervised Geometric Distance for Evaluating Real-world Noisy Point Cloud Denoising
UGD learns a Gaussian Mixture Model from clean point cloud patches to evaluate denoising degradation without requiring ground-truth clean data.