3DGS-HPC: Distractor-free 3D Gaussian Splatting with Hybrid Patch-wise Classification

Jiahao Chen, Yipeng Qin, Ganlong Zhao, Xin Li, Wenping Wang, Guanbin Li
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:16442-16460, 2026.

Abstract

3D Gaussian Splatting (3DGS) has demonstrated remarkable performance in novel view synthesis and 3D scene reconstruction, but its quality often degrades in real-world environments due to transient distractors, such as moving objects and varying shadows. Existing methods commonly introduce semantic priors from pre-trained vision models either to group pixels into coherent regions or to define perceptual error metrics. However, semantic grouping is often misaligned with the binary static/transient distinction, while perceptual features can be fragile under appearance perturbations introduced during 3DGS optimization. We propose 3DGS-HPC, a framework that addresses these issues by combining two complementary principles: a patch-wise classification strategy that leverages local spatial consistency for robust region-level decisions, and a hybrid classification metric that adaptively integrates photometric and perceptual cues for more reliable separation. Extensive experiments demonstrate the superiority and robustness of our method in mitigating distractors to improve 3DGS-based novel view synthesis. Our project page is https://cnhaox.github.io/3DGS-HPC/.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-chen26dm, title = {3{DGS}-{HPC}: Distractor-free 3{D} {G}aussian Splatting with Hybrid Patch-wise Classification}, author = {Chen, Jiahao and Qin, Yipeng and Zhao, Ganlong and Li, Xin and Wang, Wenping and Li, Guanbin}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {16442--16460}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/chen26dm/chen26dm.pdf}, url = {https://proceedings.mlr.press/v306/chen26dm.html}, abstract = {3D Gaussian Splatting (3DGS) has demonstrated remarkable performance in novel view synthesis and 3D scene reconstruction, but its quality often degrades in real-world environments due to transient distractors, such as moving objects and varying shadows. Existing methods commonly introduce semantic priors from pre-trained vision models either to group pixels into coherent regions or to define perceptual error metrics. However, semantic grouping is often misaligned with the binary static/transient distinction, while perceptual features can be fragile under appearance perturbations introduced during 3DGS optimization. We propose 3DGS-HPC, a framework that addresses these issues by combining two complementary principles: a patch-wise classification strategy that leverages local spatial consistency for robust region-level decisions, and a hybrid classification metric that adaptively integrates photometric and perceptual cues for more reliable separation. Extensive experiments demonstrate the superiority and robustness of our method in mitigating distractors to improve 3DGS-based novel view synthesis. Our project page is https://cnhaox.github.io/3DGS-HPC/.} }
Endnote
%0 Conference Paper %T 3DGS-HPC: Distractor-free 3D Gaussian Splatting with Hybrid Patch-wise Classification %A Jiahao Chen %A Yipeng Qin %A Ganlong Zhao %A Xin Li %A Wenping Wang %A Guanbin Li %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-chen26dm %I PMLR %P 16442--16460 %U https://proceedings.mlr.press/v306/chen26dm.html %V 306 %X 3D Gaussian Splatting (3DGS) has demonstrated remarkable performance in novel view synthesis and 3D scene reconstruction, but its quality often degrades in real-world environments due to transient distractors, such as moving objects and varying shadows. Existing methods commonly introduce semantic priors from pre-trained vision models either to group pixels into coherent regions or to define perceptual error metrics. However, semantic grouping is often misaligned with the binary static/transient distinction, while perceptual features can be fragile under appearance perturbations introduced during 3DGS optimization. We propose 3DGS-HPC, a framework that addresses these issues by combining two complementary principles: a patch-wise classification strategy that leverages local spatial consistency for robust region-level decisions, and a hybrid classification metric that adaptively integrates photometric and perceptual cues for more reliable separation. Extensive experiments demonstrate the superiority and robustness of our method in mitigating distractors to improve 3DGS-based novel view synthesis. Our project page is https://cnhaox.github.io/3DGS-HPC/.
APA
Chen, J., Qin, Y., Zhao, G., Li, X., Wang, W. & Li, G.. (2026). 3DGS-HPC: Distractor-free 3D Gaussian Splatting with Hybrid Patch-wise Classification. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:16442-16460 Available from https://proceedings.mlr.press/v306/chen26dm.html.

Related Material