MSP-MVS: 多粒度分割先验引导的多视图立体
MSP-MVS: Multi-Granularity Segmentation Prior Guided Multi-View Stereo
AI总结：
MSP-MVS通过引入多粒度分割先验，解决多视图立体中边缘跳过导致的变形不稳定性，提升匹配精度与泛化能力。
摘要：
最近，基于补丁变形的方法在多视图立体中展现出显著的优势，通过自适应扩展补丁的接收场来帮助重建无纹理区域。然而，此类方法主要集中在构造变形补丁时寻找无匹配歧义的像素（即可靠像素），而忽视了由意外边缘跳过引起的变形不稳定性，导致潜在的匹配扭曲。针对此问题，我们提出MSP-MVS方法，引入多粒度分割先验以指导边缘受限的补丁变形。具体而言，为了避免意外边缘跳过，我们首先聚合并进一步细化从Semantic-SAM获得的多粒度深度边缘作为先验，以引导补丁变形在深度连续（即均匀）区域。此外，为了解决由边缘受限补丁变形引起的注意力不平衡问题，我们实现了自适应等分布置和相关可靠像素（即锚点）的拆分-聚类，从而促进注意力一致的补丁变形。最后，为了防止由固定采样模式引起的变形补丁陷入局部最小匹配成本，我们引入了视差采样协同的3D优化，以帮助识别全局最小匹配成本。在ETH3D和Tanks & Temples基准测试中，我们的方法证明取得了最先进的性能，并具有显著的泛化能力。
英文摘要：
Recently, patch deformation-based methods have demonstrated significant strength in multi-view stereo by adaptively expanding the reception field of patches to help reconstruct textureless areas. However, such methods mainly concentrate on searching for pixels without matching ambiguity (i.e., reliable pixels) when constructing deformed patches, while neglecting the deformation instability caused by unexpected edge-skipping, resulting in potential matching distortions. Addressing this, we propose MSP-MVS, a method introducing multi-granularity segmentation prior for edge-confined patch deformation. Specifically, to avoid unexpected edge-skipping, we first aggregate and further refine multi-granularity depth edges gained from Semantic-SAM as prior to guide patch deformation within depth-continuous (i.e., homogeneous) areas. Moreover, to address attention imbalance caused by edge-confined patch deformation, we implement adaptive equidistribution and disassemble-clustering of correlative reliable pixels (i.e., anchors), thereby promoting attention-consistent patch deformation. Finally, to prevent deformed patches from falling into local-minimum matching costs caused by the fixed sampling pattern, we introduce disparity-sampling synergistic 3D optimization to help identify global-minimum matching costs. Evaluations on ETH3D and Tanks & Temples benchmarks prove our method obtains state-of-the-art performance with remarkable generalization.
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。
