SGAD-SLAM: Splatting Gaussians at Adjusted Depth for Better Radiance Fields in RGBD SLAM

CVPR 2026

Pengchong Hu      Zhizhong Han
Machine Perception Lab, Wayne State University, Detroit, USA
Paper Code

Abstract

3D Gaussian Splatting (3DGS) has made huge progress in RGBD SLAM. Current methods usually use 3D Gaussians or view-tied 3D Gaussians to represent radiance fields in tracking and mapping. However, these Gaussians are either too flexible or too limited in movements, resulting in slow convergence or limited rendering quality. To resolve this issue, we adopt pixel-aligned Gaussians but allow each Gaussian to adjust its position along its ray to maximize the rendering quality, even if Gaussians are simplified for improving scalability. To speed up the tracking, we model the depth distribution around each pixel as a Gaussian function, and then use these points to align each frame to the 3D scene quickly. We report our evaluations on widely used benchmarks, justify our designs, and show advantages over the latest methods in view rendering, camera tracking, runtime, and storage complexity.

Method

Overview of our method. (a–e) Robust tracking with geometry Gaussians. We initialize the \(i\)-th frame's camera pose \(p_i\) by using the constant speed assumption if no explicit illustration is provided (a) and optimize the pose \(p_i\) by aligning depth to the scene using geometric similarity (b–c). New local Gaussians \(T_i\) are then merged into the global map \(T\) for continuous tracking (d). (f) Efficient mapping with appearance Gaussians. To improve scalability and rendering quality, we learn appearance Gaussians by progressively refining pixel-wise depth offsets, demonstrating clear improvements in rendering performance across iterations. Gaussian centers are shown for visual clarity.

Video

Visual Comparisons

Ours
NICE-SLAM
Ours
SplaTAM
Ours
GS-ICP SLAM
Ours
VTGaussian-SLAM

Acknowledgement

This project was partially supported by an NVIDIA academic award and a Richard Barber research award.

BibTeX

@InProceedings{Hu2026sgadslam,
                title = {SGAD-SLAM: Splatting Gaussians at Adjusted Depth for Better Radiance Fields in RGBD SLAM},
                author = {Hu, Pengchong and Han, Zhizhong},
                booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
                year = {2026}
                }