Splat2Real：基于物理AI的新型视角缩放与3D高斯点云技术
Splat2Real: Novel-view Scaling for Physical AI with 3D Gaussian Splatting
AI总结：
Splat2Real通过CN-Coverage策略提升物理AI在新型视角下的鲁棒性，实现更稳定的深度预训练和下游控制代理性能。
摘要：
物理AI在训练和部署之间面临视角变化问题，新型视角的鲁棒性对于单目RGB到3D感知至关重要。我们将Real2Render2Real单目深度预训练视为一种模仿学习风格的监督，从数字双胞胎 oracle 获取指导：学生深度网络模仿专家度量深度/可见性，由场景网格渲染产生，而3DGS提供可扩展的新型视角观察。我们提出了Splat2Real，核心在于新型视角缩放：性能更多取决于添加的视角而非原始视角数量。我们引入了CN-Coverage，一种结合覆盖和新颖性的课程学习策略，通过几何增益和外推惩罚贪婪选择视角，并引入质量感知的守护栏回退机制以应对低可靠性教师。在20个TUM RGB-D序列中，使用步调匹配的预算（N=0到2000个额外渲染视角，N唯一<=500且为较大预算进行重采样），朴素缩放不稳定；CN-Coverage在相对Robot/Coverage策略下缓解了最坏情况的回归，而GOL-Gated CN-Coverage在中等-高预算下提供了最稳定的性能，同时具有最低的高新颖性尾部误差。下游控制代理结果与N的比较通过在视角变化下转移安全/进展权衡，提供了具身相关性的证据。
英文摘要：
Physical AI faces viewpoint shift between training and deployment, and novel-view robustness is essential for monocular RGB-to-3D perception. We cast Real2Render2Real monocular depth pretraining as imitation-learning-style supervision from a digital twin oracle: a student depth network imitates expert metric depth/visibility rendered from a scene mesh, while 3DGS supplies scalable novel-view observations. We present Splat2Real, centered on novel-view scaling: performance depends more on which views are added than on raw view count. We introduce CN-Coverage, a coverage+novelty curriculum that greedily selects views by geometry gain and an extrapolation penalty, plus a quality-aware guardrail fallback for low-reliability teachers. Across 20 TUM RGB-D sequences with step-matched budgets (N=0 to 2000 additional rendered views, with N unique <= 500 and resampling for larger budgets), naive scaling is unstable; CN-Coverage mitigates worst-case regressions relative to Robot/Coverage policies, and GOL-Gated CN-Coverage provides the strongest medium-high-budget stability with the lowest high-novelty tail error. Downstream control-proxy results versus N provides embodied-relevance evidence by shifting safety/progress trade-offs under viewpoint shift.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
