SGSoft：通过模板引导软信号学习融合语义-几何特征的三维形状对应
SGSoft: Learning Fused Semantic-Geometric Features for 3D Shape Correspondence via Template-Guided Soft Signals
在规范模板上构建测地线对应场，并用预训练语义先验监督学习稠密多模态描述子，从而在描述子空间做最近邻检索、单次前馈即可得到拓扑不变的三维形状对应；这套「模板+软信号监督+描述子检索」的范式可直接迁移到非刚性配准、形变迁移与跨类别形状检索。
原文依据 learns multimodal dense descriptors guided by pretrained semantic priors with this geodesic correspondence field supervision
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可形变三维形状之间学习稠密对应仍是一项长期挑战，其原因在于结构差异、非等距形变以及拓扑不一致。现有方法通常需要在泛化性、几何保真度与效率之间做出取舍。为此，我们提出 SGSoft，一个统一的內蕴流程：(i) 在规范模板上构建测地线对应场；(ii) 在该测地线对应场的监督下，学习由预训练语义先验引导的多模态稠密描述子；(iii) 通过在描述子空间中进行最近邻搜索，以单次前馈完成稠密对应检索。该形式化方法在大幅姿态变化、结构差异与重网格化条件下实现了稳定且拓扑不变的监督。SGSoft 在取得最优的跨类别泛化的同时，也在已有方法中提供了最佳的精度-效率权衡。它无需预对齐、成对优化或后处理精化，即可实现接近实时的推理。所学习的描述子可有效迁移至语义分割与形变迁移等下游任务，为稠密三维对应建立了一种可扩展且可直接部署的范式。
英文原文摘要 1247 字符
Learning dense correspondences across deformable 3D shapes remains a long-standing challenge due to structural variability, non-isometric deformation, and inconsistent topology. Existing methods typically trade off generalization, geometric fidelity, and efficiency. We address this by proposing SGSoft, a unified intrinsic pipeline that (i) constructs a geodesic correspondence field on a canonical template, (ii) learns multimodal dense descriptors guided by pretrained semantic priors with this geodesic correspondence field supervision, (iii) retrieves dense correspondences in a single feed-forward pass via nearest-neighbor search in descriptor space. This formulation enables stable and topology-invariant supervision under large pose variation, structural differences, and remeshing. SGSoft achieves state-of-the-art inter-category generalization while offering the best accuracy-efficiency trade-off among prior methods. It also achieves near real-time inference without pre-alignment, pairwise optimization, or post-refinement. Learned descriptors can be transferred effectively to downstream tasks such as semantic segmentation and deformation transfer, establishing a scalable and deployment-ready paradigm for dense 3D correspondence.