- Rating
- 3
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 3
Read review
Paper summary
This paper studies how to recover a clean 3D scene from images taken through smoke that changes during capture. VeilSplat jointly fits a Gaussian scene and a separate smoke model, accounting for attenuation and scattering while preserving the scene’s occlusion weights. The fitted scene can then be rendered without the smoke model. On five RealX3D scenes, it improves clean novel-view quality over the five evaluated baselines. The ablations examine coverage separation and changing smoke density, while a separate experiment evaluates forward scattering by fixing the scene and measuring the fit to smoky observations.
Critical strengths
-
The improvement on the evaluated real scenes is substantial and consistent. VeilSplat gains 5.44 dB in mean PSNR over the strongest evaluated baseline and has the highest PSNR and SSIM in each of the five scenes (Tables 1 and 3). The visual comparisons also show clearer boundaries and surface detail. These results are measured at held-out clean viewpoints, so they provide direct evidence that the method improves scene recovery.
-
The varying smoke model has a measurable benefit despite its small size. Adding two shared spatial bases and per-capture coefficients improves mean PSNR by 1.10 dB over the static density field, with better PSNR and LPIPS in all five scenes under matched training settings (Table 4). This supports the decision to model changes in smoke across captures. The extra model is only needed during fitting; clean views still use ordinary Gaussian rendering.
Critical weaknesses
The contribution of forward scattering to clean scene recovery is not directly evaluated. The ablation in Section 4.5 and Appendix G freezes an already fitted Gaussian scene, refits only the medium, and compares fitting errors on the smoky training images. This supports the authors’ claim that the branch improves observation fitting, but does not establish whether it improves clean novel-view quality when the scene and medium are jointly optimized. Since forward scattering is an important part of the method, this gap limits my assessment of its contribution, though it is not sufficient to undermine the experimental results for the method as a whole.
Questions for authors
-
How exactly is the no-forward-scattering condition constructed? Which model was used to obtain the frozen Gaussian scene? How are ω and the airlight weight handled when the forward branch is removed? These details matter for interpreting the error differences in Figure 7.
-
DehazeGS also reconstructs clear novel views from multi-view foggy RGB images and is discussed in the paper. Could you clarify why it was not included in the experimental comparison? If a usable implementation is available, a comparison on the same five RealX3D scenes, with its pretrained depth prior retained and clearly noted, would help assess VeilSplat relative to this closely related method. If such a comparison is not feasible, please explain the implementation or evaluation constraints.
2 comments
Thank you for the reply. I have no major concerns and maintain my recommendation. And I have updated the AI-use documentation field.
Response to Reviewer 5FYA
Response to Reviewer 5FYA
Thank you for your positive, constructive review and for recognizing our clean novel-view gains and compact time-varying smoke representation.
Response: Jointly optimizing the 3D scene and smoke model end-to-end (withholding smoky captures 4, 10, 16, and 22 per scene), physical forward and scene-lit weights lift clean test PSNR by +0.35 dB for the full model, and covariance dilation achieves the best SSIM (0.577) and LPIPS (0.531) (Section 4.6, Tables 5, S13).
We jointly train three variants on Hinoki, Koharu, and Shirohana (18, 21, and 22 smoky training captures): (1) Full model (using the physical small-angle weight t1−f−t, which matches fτ at small τ and vanishes as t→0; Section 3.2, Figure 2a), (2) No forward lobe (f=ω=0, standard airlight (1−t)A), and (3) No dilation (h→0), keeping f,ω active without covariance broadening:
Response: Both arms started from the converged scene of the f=0 variant of Eq. (1) and optimized only the smoke parameters for 500 steps; the no-forward arm dropped the dilated pass while keeping (1−ω)(1−t)A with ω learnable (Section S16; Figure 7 of the original paper is now Figure S8).
Response: DehazeGS (Yu et al., 2026) has no public code or checkpoints (checked November 2026), so we retrained two restore-then-reconstruct baselines (DCP → 3DGS and pretrained DehazeFormer → 3DGS) under one protocol, where VeilSplat leads on all metrics (+0.97 dB over DehazeFormer → 3DGS and +2.84 dB over DCP → 3DGS; +5.17 dB over the strongest official RealX3D baseline in Table 2) and stays at or above 13.45 dB on every scene (Section S4, Table S3, Figure 3).
VeilSplat reaches ≥13.45 dB clean PSNR on every room, whereas
DCP → 3DGSdrops to 11.11 dB on Shirohana andDehazeFormer → 3DGSto 11.31 dB on Hinoki (Table S3). If an official DehazeGS implementation becomes public during discussion, we will evaluate it on all five scenes with its pretrained depth prior preserved and noted.Thank you again for a clear, focused review that strengthened the joint-training forward-lobe evaluation and restoration comparisons of the paper.