发表机构
University of Chicago; University of Southern California; Technion（芝加哥大学; 南加州大学; 以色列理工学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
RADmesh是一种带重网格化的网格变形方法，可实现大变形、抗噪声，能生成适配几何的规整各向同性网格，在局部和全局变形任务上表现优于现有方法。
AI 中文摘要
我们提出了一种增强重网格化的方法，用于在视觉损失下生成式变形形状。直观来看，若不对网格进行重新三角剖分，对其进行足够剧烈的变形极易损害单元质量，即便这类大的几何变化在语义上是所需的。因此，形状变形方法可从改变三角剖分中获益；然而，大多数基于文本、视觉监督的生成式网格变形方法并未采用重网格化。重网格化是一种离散操作，已被证明与视觉损失提供的、以噪声著称的监督信号结合极具挑战性。我们提出了一种基于顶点的变形优化量，能实现大变形且对这类噪声具有鲁棒性；我们使用各向同性重网格化器定期重网格化，该重网格化器会插值并传递变形优化状态。这使得在从粗到细增加分辨率的过程中，能够实现连续的、几何信息驱动的进展。所得形状的三角剖分与其优化后的几何结构相适配，且具有规整的各向同性单元。此外，我们的方法是可定位的，能够在基础形状上生成具有表达性细节的新特征，同时保持其余部分不变。我们在多种形状和提示（包括局部和全局变形）上展示了该方法的有效性，并证明了其优越的视觉质量和三角形效率。我们的项目页面位于此https URL。
英文摘要
We propose a remeshing-enhanced method for generatively deforming shapes with visual losses. It is intuitive that sufficiently drastic deformations of a mesh without changing its triangulation can easily compromise element quality, even if such large geometry changes may be semantically desired. Shape deformation methods could thus benefit from changing the triangulation; however, this is not done by most generative, text-based, visually-supervised mesh deformation methods. Remeshing is a discrete operation, proven to be especially challenging to couple with the notoriously noisy supervision signal provided by visual losses. We propose a vertex-based deformation optimization quantity capable of large deformations and robustness to such noise; we periodically remesh using an isotropic remesher that interpolates and carries forward the deformation optimization state. This enables continuous, geometry-informed progress in coarse-to-fine addition of resolution. The resulting shapes' triangulations fit their optimized geometry and have neat isotropic elements. Further, our method is localizable, able to grow new features on a base shape with expressive detail, leaving the rest unchanged. We showcase the effectiveness of our method on a variety of shapes and prompts, both local and global deformations, and demonstrate its superior visual quality and triangle efficiency. Our project page is at https://threedle.github.io/radmesh.
CommentsECCV 2026 (Oral). Our project page is at https://threedle.github.io/radmesh
