从无约束图像中泛化的稀疏视角三维重建
Generalizable Sparse-View 3D Reconstruction from Unconstrained Images
提出无需逐场景优化的前馈式稀疏视角室外重建框架，利用学习到的几何先验从无位姿互联网图像直接预测深度、相机参数和规范空间三维高斯，并通过外观适配器与语义分割处理光照变化和瞬态遮挡，为稀疏输入下的实时新视角合成提供了可复用的技术路线。
原文依据 We present GenWildSplat, a feed-forward framework for sparse-view outdoor reconstruction that requires no per-scene optimization.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论在光照变化和瞬态遮挡的真实条件下，从稀疏、无位姿图像重建三维场景仍然具有挑战性。现有方法依赖带有外观嵌入或动态掩码的逐场景优化，需要大量逐场景训练，并且在稀疏视角下失效。此外，在有限场景上的评估引发了对泛化性的质疑。我们提出 GenWildSplat，一个用于稀疏视角室外重建的前馈框架，无需逐场景优化。给定无位姿的互联网图像，GenWildSplat 利用学习到的几何先验，在规范空间中预测深度、相机参数和三维高斯。外观适配器为目标光照条件调制外观，而语义分割处理瞬态物体。通过在合成数据和真实数据上进行课程学习，GenWildSplat 能够泛化到多样的光照和遮挡模式。在 PhotoTourism 和新的 20 场景 MegaScenes 基准上的评估表明，其前馈重建质量达到最优，实现了无需测试时优化的实时推理。
英文原文摘要 1137 字符
Reconstructing 3D scenes from sparse, unposed images remains challenging under real-world conditions with varying illumination and transient occlusions. Existing methods rely on scene-specific optimization with appearance embeddings or dynamic masks, requiring extensive per-scene training and failing under sparse views. Moreover, evaluations on limited scenes raise questions about generalization. We present GenWildSplat, a feed-forward framework for sparse-view outdoor reconstruction that requires no per-scene optimization. Given unposed internet images, GenWildSplat predicts depth, camera parameters, and 3D Gaussians in a canonical space using learned geometric priors. An appearance adapter modulates appearance for target lighting conditions, while semantic segmentation handles transient objects. Through curriculum learning on synthetic and real data, GenWildSplat generalizes across diverse illumination and occlusion patterns. Evaluations on PhotoTourism and a new 20-scene MegaScenes benchmark demonstrate state-of-the-art feed-forward reconstruction quality, achieving real-time inference without test-time optimization.