写在关键处：面向 3D 高斯泼溅的策略引导水印
Write Where It Matters: Policy-Guided Watermarks for 3D Gaussian Splatting
把水印嵌入位置与强度建模为马尔可夫决策过程，用轻量策略网络按即时奖励逐锚点调节梯度尺度；做 3DGS 版权保护或需要在不破坏渲染质量的前提下向高斯参数注入可控扰动时，这套策略引导的逐锚点梯度缩放可直接复用。
原文依据 W2M uses a lightweight policy network to allocate precise Gaussian updates directly from immediate reward feedback
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期 3D 高斯泼溅（3DGS）的进展实现了照片级真实感的实时渲染，但也加剧了未经授权复制与再分发的风险。现有的 3DGS 水印方法通常依赖手工设定的阈值或全局固定的超参数来平衡不可见性与鲁棒性，使其嵌入行为是静态的、与场景无关的。我们转而将 3DGS 水印建模为目标导向的决策过程，并提出写在关键处（W2M），这是首个基于强化学习、能够自适应地学习在何处嵌入以及嵌入多少的框架。通过将嵌入过程建模为马尔可夫决策过程，W2M 使用轻量策略网络，直接依据即时奖励反馈迭代地分配精确的高斯更新。该奖励同时激励渲染空间的不可见性以及在各类图像级与模型级失真下的解码鲁棒性。为提升效率，W2M 在围绕可学习锚点组织的结构化 3DGS 骨干网络上运行，并施加策略引导的逐锚点梯度缩放。在 Blender、LLFF 与 Mip-NeRF 360 数据集上的大量实验表明，W2M 在标准与对抗条件下均取得了最优的比特准确率、较强的感知保真度与结构一致性。
英文原文摘要 1309 字符
Recent advances in 3D Gaussian Splatting (3DGS) enable photorealistic real-time rendering but also increase the risks of unauthorized copying and redistribution. Existing 3DGS watermarking methods typically rely on handcrafted thresholds or globally fixed hyperparameters to balance invisibility and robustness, making their embedding behavior static and scene-agnostic. We instead formulate 3DGS watermarking as a goal-directed decision process and introduce Write Where It Matters (W2M), the first reinforcement learning-based framework that adaptively learns where and how much to embed. By modeling the embedding process as a Markov Decision Process, W2M uses a lightweight policy network to allocate precise Gaussian updates directly from immediate reward feedback, iteratively. The reward incentivizes both rendering-space invisibility and decoding robustness under various image- and model-level distortions. To achieve efficiency, W2M operates on a structured 3DGS backbone organized around learnable anchors and applies policy-guided per-anchor gradient scaling. Extensive experiments across the Blender, LLFF, and Mip-NeRF 360 datasets demonstrate that W2M achieves state-of-the-art bit accuracy, strong perceptual fidelity, and structural consistency under both standard and adversarial conditions.