- Rating
- 3
- Confidence
- 2
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 2
Read review
Paper summary
This paper proposes a new 3DGS poisoning attack which targets post-reconstruction tampering. The attacker leaves the original benign Gaussian scene unchanged, and injects a small set of new Gaussians. The goal is to render a target pattern at a chosen victim viewpoint, while keeping the benign viewpoints stay close to the original clean appearance, without introducing geometric artifacts. Experiments cover 3 datasets, and the attack exhibit a strong attack-view PSNR and faster optimization time compared with existing reconstruction phase attacks.
Critical strengths
-
The paper identifies a research gap of post-reconstruction tampering. This new setting introduces less optimization time compared with previous reconstruction stage attacks.
-
The proposed method is elegant and mathematically sound.
-
The evaluation is broad on three datasets and 24 scenes. The authors introduces semantic metrics (CLIP and YOLO) which go beyond PSNR.
Critical weaknesses
- The first-order approximation is not well-motivated: The paper's narrative is that low-opacity Gaussians make the problem "first-order". This is supposed to enable decoupled and cached rendering. However, caching does not require the approximation.
Questions for authors
-
Can you compare exact cached compositing (prefix product) against the linearized surrogate, with both evaluated by the true renderer? Does the linearization improve anything?
-
How robust is the attack to opacity-threshold pruning (e.g., removing ) and to detectors based on scale/opacity statistics?
5 comments
Following up questions
Thank you for your response. Regarding the response, I have two follow-up questions:
-
Can you show the value distribution of exact and the approximated ? For example, what is the mean standard deviation of these two values, across both attacked views and benign views? What would the distribution of their difference be? How would that change according to the number of injected adversarial Gaussians?
-
I can understand that such first-order approximation can make optimization faster and thus reduce attack time. But why approximation will make the optimization results even better than the exact computation on the attacked views? Why is it not the case for benign views? Could you explain the details?
Response to Follow-up Q1. We appreciate your follow-up and the opportunity to address the remaining concerns. Below, we report the mean and standard deviation of and its first-order approximation , evaluated on our Earth image attack presented in Appendix D.10, which includes one attack view and 145 benign views.
| View | Exact | Approx. | |
|---|---|---|---|
| Attack | 0.861 0.128 | 0.840 0.170 | 0.021 0.047 |
| Benign | 0.774 0.210 | 0.694 0.368 | 0.079 0.174 |
Overall, the first-order approximation underestimates the exact transmittance, with a larger discrepancy at benign viewpoints than at the attack viewpoint. To see how changes, we average it across rays grouped by the number of injected Gaussians along each ray.
| # Injected Gaussians Number | Attack View Avg | Benign Views Avg |
|---|---|---|
| 1–2 | 0 | 0 |
| 5 | 0.0019 | 0.0018 |
| 10 | 0.011 | 0.011 |
| 20 | 0.070 | 0.057 |
| 30 | 0.147 | 0.111 |
As more Gaussians overlap along a ray, the omitted higher-order opacity interactions accumulate, increasing the approximation error.
Response to Follow-up Q2. The adversarial optimization, just as 3DGS, is nonlinear and nonconvex, so the first-order approximation may outperform exact computation. Moreover, our objective maximizes attack-view fidelity while keeping benign-view degradation within a small bound. Thus, greater benign-view degradation may coexist with improved attack-view fidelity. To further investigate how approximation errors contribute to these performance differences, we compare the gradients obtained using the first-order approximation and exact computation on the Earth image attack:
| View | Gradient Cosine Similarity ↑ | Sign Disagreement ↓ |
|---|---|---|
| Attack view | 0.994 | 0.38% |
| Benign views (per-view) | 0.951 0.074 | 3.20% |
| Benign views (aggregated) | 0.993 | 0.77% |
We observe that gradients obtained using the first-order approximation align more closely with exact gradients at the attack view. Despite strong alignment when benign-view gradient contributions are aggregated (0.993 cosine similarity and 0.77% sign disagreement), individual benign-view gradients show weaker alignment (0.951 cosine similarity and 3.2% sign disagreement). This suggests that the approximation is more likely to misrepresent gradient directions for individual benign views, potentially leading to greater fidelity degradation.
We thank the reviewer for the constructive comments and positive feedback on our paper. Regarding the concerns of the reviewer 8XnC, we provide the following responses.
Response to W1. We thank the reviewer for this observation. We would like to clarify the role of first-order approximation in our decoupled rendering formulation. Our caching mechanism does not involve any approximation. Specifically, we precompute and cache the exact contributions of the frozen original Gaussians, including the original-scene color Cbase, transmittance Tibase, and suffix color contribution Γi.
The first-order approximation is introduced during subsequent adversarial compositing, where we approximate the transmittance induced by the injected low-opacity Gaussians:
\approx 1-\sum_{\substack{j<i\\\j\in\mathcal I_{\mathrm{inj}}}}\alpha_j =1-S_i.$$ This leads to our simplified decoupled rendering formulation: $$\widetilde C_p=C_{\mathrm{base}}+ \sum_{i\in\mathcal I_{\mathrm{inj}}} \left[ T_i^{\mathrm{base}}(1-S_i)\alpha_i c_i -\alpha_i\Gamma_i \right].$$ Therefore, **first-order approximation is part of our decoupled rendering formulation, but is not required for caching itself**. Its purpose is to simplify adversarial transmittance interactions and gradient computation during optimization. We also acknowledge that the phrase *"cached first-order approximation"* in our ablation description may have conflated these two operations. We will revise this wording to explicitly distinguish exact caching of the original scene from first-order approximation of adversarial compositing. **Response to Q1.** As clarified in our response to W1, **we do not use first-order approximation during caching**. The contributions of the frozen original Gaussians are computed exactly. The first-order approximation is introduced only when computing the transmittance effects of the injected adversarial Gaussians during optimization. We believe that this approximation can facilitate optimization by reducing nonlinear interactions among injected Gaussians. Specifically, exact transmittance involves multiplicative dependencies, $$U_i=\prod_{j<i}(1-\alpha_j),$$ whereas our first-order approximation simplifies them into an additive form, $$\widetilde U_i=1-\sum_{j<i}\alpha_j.$$ This simplification reduces higher-order interactions and may alleviate gradient attenuation, potentially improving optimization stability and convergence. To further investigate this effect, we extend our existing ablation by introducing Decoupled Opt (Exact), which uses exact adversarial compositing with prefix products, alongside the original Decoupled Opt, which uses our first-order approximation. Both settings are identical except for the compositing strategy, and all results are evaluated using the full 3DGS renderer. This comparison isolates the contribution of first-order approximation to optimization efficiency and attack performance. | Method | Attack PSNR ↑ | Benign PSNR ↑ | Time (min) ↓ | |:---|---:|---:|---:| | Decoupled Opt (Exact) | 32.12 | 28.78 | 34.89 | | Decoupled Opt (Ours) | 34.42 | 28.47 | 24.28 | With exact compositing, optimization becomes more challenging due to the multiplicative transmittance interactions among injected Gaussians. As a result, optimization may fail for some viewpoints selected by the outer loop, leading to additional viewpoint searches and longer overall optimization time. The increased optimization difficulty can still reduce attack-view quality or require more viewpoint selection attempts to achieve a successful attack. **Response to Q2.** We additionally evaluate opacity-based pruning ($\alpha < 0.05$) on the original Bicycle scene, with visual comparisons shown in **Appendix Fig. 12.** At this threshold, 27.69% (1,333,851/4,816,275) of the original Gaussians are removed. This introduces visible rendering differences, with SSIM dropping below 0.95, PSNR dropping below 30 dB. We also test a lower pruning threshold of $\alpha < 0.01$ and find that the attack remains largely unaffected, with attack-view PSNR still above 30 dB. These results highlight the trade-off between attack removal and scene fidelity. These results show that aggressive pruning can remove low-opacity adversarial Gaussians, but also affects legitimate scene content. Moreover, our injected Gaussians account for less than 1% of all Gaussians below this opacity threshold, making them difficult to distinguish from legitimate Gaussians using opacity statistics alone. Standard 3DGS implementations also use much lower pruning thresholds (e.g., 0.005), and downstream users do not know the attacker's opacity bound in advance.