Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2601.16736 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914541502201856 |
|---|---|
| author | Ding, Renjie Wang, Yaonan Liu, Min Zhu, Jialin Wang, Jiazheng Zhao, Jiahao Shen, Wenting He, Feixiang Chen, Xiang |
| author_facet | Ding, Renjie Wang, Yaonan Liu, Min Zhu, Jialin Wang, Jiazheng Zhao, Jiahao Shen, Wenting He, Feixiang Chen, Xiang |
| contents | 3D Gaussian Splatting (3DGS) has emerged as a powerful technique for real-time novel view synthesis. As an explicit representation optimized through gradient propagation among primitives, optimization widely accepted in deep neural networks (DNNs) is actually adopted in 3DGS, such as synchronous weight updating and Adam with the adaptive gradient. However, considering the physical significance and specific design in 3DGS, there are two overlooked details in the optimization of 3DGS: (i) update step coupling, which induces optimizer state rescaling and costly attribute updates outside the viewpoints, and (ii) gradient coupling in the moment, which may lead to under- or over-effective regularization. Nevertheless, such a complex coupling is under-explored. After revisiting the optimization of 3DGS, we take a step to decouple it and recompose the process into: Sparse Adam, Re-State Regularization and Decoupled Attribute Regularization. Taking a large number of experiments under the 3DGS and 3DGS-MCMC frameworks, our work provides a deeper understanding of these components. Finally, based on the empirical analysis, we re-design the optimization and propose AdamW-GS by re-coupling the beneficial components, under which better optimization efficiency and representation effectiveness are achieved simultaneously. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_16736 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | A Step to Decouple Optimization in 3DGS Ding, Renjie Wang, Yaonan Liu, Min Zhu, Jialin Wang, Jiazheng Zhao, Jiahao Shen, Wenting He, Feixiang Chen, Xiang Computer Vision and Pattern Recognition 3D Gaussian Splatting (3DGS) has emerged as a powerful technique for real-time novel view synthesis. As an explicit representation optimized through gradient propagation among primitives, optimization widely accepted in deep neural networks (DNNs) is actually adopted in 3DGS, such as synchronous weight updating and Adam with the adaptive gradient. However, considering the physical significance and specific design in 3DGS, there are two overlooked details in the optimization of 3DGS: (i) update step coupling, which induces optimizer state rescaling and costly attribute updates outside the viewpoints, and (ii) gradient coupling in the moment, which may lead to under- or over-effective regularization. Nevertheless, such a complex coupling is under-explored. After revisiting the optimization of 3DGS, we take a step to decouple it and recompose the process into: Sparse Adam, Re-State Regularization and Decoupled Attribute Regularization. Taking a large number of experiments under the 3DGS and 3DGS-MCMC frameworks, our work provides a deeper understanding of these components. Finally, based on the empirical analysis, we re-design the optimization and propose AdamW-GS by re-coupling the beneficial components, under which better optimization efficiency and representation effectiveness are achieved simultaneously. |
| title | A Step to Decouple Optimization in 3DGS |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2601.16736 |