从稀疏全景图重建3D场景，无需传统SfM/SLAM。
PanoImager: Geometry-Guided Novel View Synthesis and Reconstruction from Sparse Panoramic Views
- 用前馈姿态与深度先验，结合几何引导的扩散模型补全视图。
- 在极端稀疏条件下仍保持稳定，跨视角一致性显著提升。
- 适合SfM/SLAM失败时的离线地图优化场景。
全景感知提供广域视野，但在以旋转为主、弱视差的运动下，从稀疏全景图进行3D重建仍具挑战性，此时传统SfM/SLAM初始化常因病态而不可靠。本文提出PanoImager，一种免于SfM的框架，融合前馈姿态/深度先验、几何条件化扩散视图补全与深度引导的3DGS优化。仅需少量全景图像，PanoImager将全景图分解为局部透视视图，合成辅助观测以丰富稀疏证据，并稳定高斯优化，提升跨视角一致性。多个基准测试表明，在极端稀疏条件下仍表现更优，适用于SfM/SLAM无法初始化时的离线/后台地图精修。
原文摘要 · Abstract (English)
Panoramic sensing offers wide field-of-view coverage, yet 3D reconstruction from sparse panoramas remains challenging under rotation-dominant, weak-parallax motion. In such regimes, SfM/SLAM initialization is often ill-conditioned and unreliable. We present PanoImager, an SfM-free framework that combines feed-forward pose/depth priors, geometry-conditioned diffusion view completion, and depth-guided 3DGS optimization. Given only a few panoramic images, PanoImager decomposes them into local perspective views, synthesizes auxiliary observations to enrich sparse evidence, and stabilizes Gaussian optimization for improved cross-view consistency. Experiments on multiple benchmarks show improved stability under extreme sparsity, suggesting PanoImager as an offline/background component for map refinement when SfM/SLAM fails to initialize.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。