Online 3D Gaussian mapping enables fast and high-fidelity novel view synthesis, with broad applications in augmented reality, robotics, and autonomous systems. However, many current methods assume static scenes or filter out dynamics as outliers. In this paper, we study in-view scene changes where objects are manipulated in front of the camera within a single online sequence, unlike prior works that detect changes outside the camera view or between separate captures. This setting entangles transient and persistent dynamics, as objects can be distractors while moving and persistent once they settle. While recent studies start to address distractors by removing them for clean static background, in-view scene changes remain underexplored. To resolve this, we present SettleSplat, a monocular mapping system that maintains up-to-date 3D Gaussian maps under such changes. Specifically, we lift coarse per-pixel uncertainty into Gaussian dynamicity field, then prune transient ones. For persistent changes, we derive geometric contradictions weighted by parameter-free rendered depth uncertainty to separate true changes from under-optimized regions. On Bonn and Co-Fusion, SettleSplat achieves consistent 3D Gaussian maps, outperforming current baselines on both full images and change regions while maintaining comparable speed.
Figures & tables
Figure 2: System overview. Given monocular RGB frames (a), SettleSplat estimates depth, then tracks and maps with the DROID-W backbone (b) while learning per-Gaussian dynamicity field δ that prunes transient distractors. During refinement (c), geometry contradiction score g , derived from the rendered map and estimated depth and weighted by the rendered depth uncertainty, is integrated with the tracking uncertainty into reliability map r . Frames before the settlement frame K are weighted by W=r2 , while later frames supervise the settled scene without masking.
Method
Metric
moving
moving2
placing
placing2
placing3
removing
removing2
Avg.
Full Image
SplaTAM [ 8 ]
PSNR ↑
21.38
22.84
20.51
18.74
22.75
21.95
19.39
21.08
SSIM ↑
0.817
0.847
0.771
0.778
0.847
0.851
0.824
0.819
LPIPS ↓
0.267
0.254
0.306
0.351
0.244
0.235
0.291
0.278
MonoGS [ 22 ]
PSNR ↑
18.89
19.88
17.09
21.64
20.91
24.01
19.37
20.26
SSIM ↑
0.696
0.745
0.640
0.794
0.777
0.852
0.790
0.756
Table 1: Novel view synthesis results on the Bonn dynamic dataset. We report PSNR, SSIM, and LPIPS on the full image (top) and within the change region (bottom). Best and second-best results are highlighted.
Full Image
Change Region
Method
Metric
place
sliding
handover
car
Avg.
place
sliding
handover
car
Avg.
SplaTAM [ 8 ]
PSNR ↑
22.42
29.08
13.91
28.98
23.60
20.89
31.59
13.85
16.99
20.83
SSIM ↑
0.844
0.848
0.292
0.955
0.735
0.791
0.876
0.225
0.614
0.626
LPIPS ↓
0.249
0.223
0.770
0.099
0.335
0.286
0.322
0.806
0.307
0.430
MonoGS [ 22 ]
PSNR ↑
15.71
28.57
20.53
32.14
24.24
14.16
29.53
20.59
20.90
21.29
SSIM ↑
0.713
0.807
0.750
0.967
0.809
0.614
0.868
0.703
0.715
0.725
Table 2: Novel view synthesis results on the CoFusion dataset, evaluated on the full image and within the change region. Best and second-best results are highlighted.
Figure 3: Qualitative comparison on Bonn (rows 1–3) and Co-Fusion (rows 4–6), covering moving, adding, and removing objects. Red boxes show the change regions. At old object locations, baselines leave ghosting artifacts, while SettleSplat reconstructs the updated scene cleanly. For added and moved objects, SettleSplat preserves appearance and detail, whereas the baselines blur or miss them. For removed objects, SettleSplat is competitive with other baselines while reducing ghosting.
Full Image
Change Region
Variant
PSNR ↑
SSIM ↑
LPIPS ↓
PSNR ↑
SSIM ↑
LPIPS ↓
w/o Vd
29.57
0.907
0.140
27.00
0.877
0.221
w/o K
27.56
0.888
0.161
24.50
0.839
0.259
w/o M
32.79
0.921
0.106
32.84
0.920
0.139
Full (Ours)
33.84
0.924
0.104
33.12
0.923
0.135
Table 3: Ablation study on CoFusion (averaged over all sequences). Best results are highlighted.
Bonn
CoFusion
Method
Track. FPS ↑
Map. /iter(s) ↓
Track. FPS ↑
Map. /iter(s) ↓
WildGS-SLAM [ 41 ]
1.7
14.7
4.0
14.0
DROID-W [ 16 ]
10.7
1.3
20.4
1.2
SettleSplat (Ours)
10.6
1.8
21.7
1.4
Table 4: Runtime comparison with dynamic-scene methods. We report the tracking speed in frames per second and the average mapping time per optimization iteration. Best results are in bold .
Symbol
Value
Description
Dynamicity pruning
τδ
0.5
Prune Gaussians with dynamicity
τn
20
Min. observations before a Gaussian is prunable
ϵ
1e−6
Geometry contradiction computation
Mapping loss weights
λpho
0.5
Photometric (RGB) vs. depth loss weight
Table S1: SettleSplat hyperparameters used in the Bonn and Co-Fusion experiments.
Dataset
Scene
thold
End
#
Bonn
moving_nonobstructing_box
476
776
61
moving_nonobstructing_box2
703
933
47
placing_nonobstructing_box
346
716
75
placing_nonobstructing_box2
394
674
57
placing_nonobstructing_box3
491
661
35
removing_nonobstructing_box
291
491
41
Table S2: Held-out novel-view frames per scene.
Figure S1 : Data preparation with change region (bounding boxes) on Bonn and CoFusion, covering moving, adding and removing objects. The rendering metrics are computed on full rendered images and cropped change regions.
Lifelong scene mapping under rigid object rearrangement remains a fundamental challenge in robotics. While 3D Gaussian Splatting (3DGS) enables high-fidelity modeling, primitive-level updates often cause persistent ghosting and slow recovery. We propose CubifyGS, an object-level mapping framework that shifts dynamic maintenance from passive re-optimization to active asset management. CubifyGS models movable instances as reusable Gaussian assets, detects object appearance and disappearance, and updates maps through asset retrieval, rigid transformation, and explicit pruning rather than reconstruction from scratch. To address geometric voids and local photometric mismatch after such edits, we further propose an event-triggered adaptive optimization strategy that focuses computation on affected regions. We validate our approach on a newly constructed high-fidelity dynamic benchmark, demonstrating that CubifyGS improves artifact suppression and maintenance efficiency over representative reproducible baselines in the evaluated object-rearrangement setting.
Bohan Ren, Dianyi Yang, Shiyang Liu +5
School of Automation, Beijing Institute of Technology, Beijing, China · 2Guangzhou Saite Intelligent Technology Co., Ltd.
Online scene change detection requires integrating incoming observations to determine which regions have changed relative to a reference scene. We update change beliefs as new images arrive using generalised Bayesian evidence fusion. However, the prior probability of change and evidence models are unknown beforehand, and numerical Bayesian inference can be costly. We address these challenges by learning offline from automatic edits to a reference 3D Gaussian Splatting (3DGS) model. These edits generate labelled training views without manual annotations. Together with unchanged reference observations, they provide supervision for learning scene-specific evidence models and estimating the effective pixel prior and response statistics. Online prediction combines current evidence with historical support, while closed-form local Bayes factors, parameterized by the learned response statistics, update an accumulated 3D Gaussian field. Object-guided refinement produces the final change masks. Our method achieves state-of-the-art accuracy on PASLCD and CCS3D, with 67.64% and 75.82% mIoU, respectively. On PASLCD, the complete pipeline runs at 17.2,FPS over complete query sequences, the fastest among the evaluated online methods. Project page: https://changruizhu96.github.io/work/lego-scd/
Reconstructing dynamic 3D scenes from monocular videos is a fundamental yet highly challenging task, as real-world motions often involve both long-term smooth transformations and short-term complex deformations. Existing methods either struggle to maintain temporal consistency or fail to capture high-frequency dynamics due to limited motion modeling capacity. In this work, we present Rigid-aware 4D Gaussian Splatting (RiGS), which simultaneously captures motions across multiple temporal scales. Specifically, RiGS introduces three types of Gaussian primitives: static, rigid, and transient, which represent static backgrounds, long-term low-frequency motions, and short-term high-frequency dynamics, respectively. An object-wise dynamic mask is proposed to aggregate long-range spatiotemporal motion information and guide the decomposition of static and dynamic regions. To jointly model motion across scales, rigid Gaussians are allowed to transition into transient Gaussians based on their temporal duration, and both are optimized under scene flow guidance, providing dense 3D motion supervision. Extensive experiments demonstrate that RiGS achieves state-of-the-art performance on novel view synthesis benchmarks. Code is available at \hyperlink{https://github.com/ladvu/RiGS}{https://github.com/ladvu/RiGS}.
Chenyu Wu, Wanhua Li, Zhu-Tian Chen +1
Harvard University · Zhejiang University · Nanyang Technological University +1