Improving Explicit Dynamic Gaussian Splatting Optimization via Update Mixture

Renjie Ding, Yaonan Wang, Min Liu, Jialin Zhu, Jiazheng Wang, Jiahao Zhao, Xiao Tan, Feixiang He, Xiang Chen
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:25152-25174, 2026.

Abstract

3D Gaussian Splatting (3DGS) enables real-time, high-fidelity view synthesis via explicit scene representations and has recently been extended to dynamic scene modeling. Despite their excellent rendering quality and interpretability, we find that explicit Dynamic GS often exhibits generalization degradation in scenes with large motion. Motivated by generalization behavior in deep neural optimization and the characteristics of Gaussian primitive optimization, we propose an update mixture strategy. This work focuses on two representative open-source explicit Dynamic GS pipelines and our approach consists of three components: (i) a space–time dependent Strictly Sparse Update with additional regularization to stabilize adaptive updates; (ii) a constant-corrected adaptive algorithm that alleviates over-scaling of primitive gradients and yields a stable mixture of adaptive and non-adaptive steps; and (iii) attributes mixing via Stochastic Attribute Averaging to mitigate frame-preference under motion disturbances. Experiments show consistent improvements and reduced generalization issues, highlighting the role of non-adaptive updates and the impact of frame-preference in explicit Dynamic GS optimization.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-ding26p, title = {Improving Explicit Dynamic {G}aussian Splatting Optimization via Update Mixture}, author = {Ding, Renjie and Wang, Yaonan and Liu, Min and Zhu, Jialin and Wang, Jiazheng and Zhao, Jiahao and Tan, Xiao and He, Feixiang and Chen, Xiang}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {25152--25174}, 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/ding26p/ding26p.pdf}, url = {https://proceedings.mlr.press/v306/ding26p.html}, abstract = {3D Gaussian Splatting (3DGS) enables real-time, high-fidelity view synthesis via explicit scene representations and has recently been extended to dynamic scene modeling. Despite their excellent rendering quality and interpretability, we find that explicit Dynamic GS often exhibits generalization degradation in scenes with large motion. Motivated by generalization behavior in deep neural optimization and the characteristics of Gaussian primitive optimization, we propose an update mixture strategy. This work focuses on two representative open-source explicit Dynamic GS pipelines and our approach consists of three components: (i) a space–time dependent Strictly Sparse Update with additional regularization to stabilize adaptive updates; (ii) a constant-corrected adaptive algorithm that alleviates over-scaling of primitive gradients and yields a stable mixture of adaptive and non-adaptive steps; and (iii) attributes mixing via Stochastic Attribute Averaging to mitigate frame-preference under motion disturbances. Experiments show consistent improvements and reduced generalization issues, highlighting the role of non-adaptive updates and the impact of frame-preference in explicit Dynamic GS optimization.} }
Endnote
%0 Conference Paper %T Improving Explicit Dynamic Gaussian Splatting Optimization via Update Mixture %A Renjie Ding %A Yaonan Wang %A Min Liu %A Jialin Zhu %A Jiazheng Wang %A Jiahao Zhao %A Xiao Tan %A Feixiang He %A Xiang Chen %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-ding26p %I PMLR %P 25152--25174 %U https://proceedings.mlr.press/v306/ding26p.html %V 306 %X 3D Gaussian Splatting (3DGS) enables real-time, high-fidelity view synthesis via explicit scene representations and has recently been extended to dynamic scene modeling. Despite their excellent rendering quality and interpretability, we find that explicit Dynamic GS often exhibits generalization degradation in scenes with large motion. Motivated by generalization behavior in deep neural optimization and the characteristics of Gaussian primitive optimization, we propose an update mixture strategy. This work focuses on two representative open-source explicit Dynamic GS pipelines and our approach consists of three components: (i) a space–time dependent Strictly Sparse Update with additional regularization to stabilize adaptive updates; (ii) a constant-corrected adaptive algorithm that alleviates over-scaling of primitive gradients and yields a stable mixture of adaptive and non-adaptive steps; and (iii) attributes mixing via Stochastic Attribute Averaging to mitigate frame-preference under motion disturbances. Experiments show consistent improvements and reduced generalization issues, highlighting the role of non-adaptive updates and the impact of frame-preference in explicit Dynamic GS optimization.
APA
Ding, R., Wang, Y., Liu, M., Zhu, J., Wang, J., Zhao, J., Tan, X., He, F. & Chen, X.. (2026). Improving Explicit Dynamic Gaussian Splatting Optimization via Update Mixture. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:25152-25174 Available from https://proceedings.mlr.press/v306/ding26p.html.

Related Material