MeshTailor: Cutting Seams via Generative Mesh Traversal

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ma, Xueqi, Yan, Xingguang, Zhang, Congyue, Huang, Hui
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914581189754880
author Ma, Xueqi
Yan, Xingguang
Zhang, Congyue
Huang, Hui
author_facet Ma, Xueqi
Yan, Xingguang
Zhang, Congyue
Huang, Hui
contents We present MeshTailor, the first mesh-native generative framework for synthesizing edge-aligned seams on 3D surfaces. Unlike prior optimization-based or extrinsic learning-based methods, MeshTailor operates directly on the mesh graph, eliminating projection artifacts and fragile snapping heuristics. We introduce ChainingSeams, a hierarchical serialization of the seam graph that orders chains from global structural cuts down to local details in a coarse-to-fine manner, and a dual-stream encoder that fuses topological and geometric context. Leveraging this hierarchical representation and dual-stream vertex embeddings, our MeshTailor Transformer utilizes an autoregressive pointer layer to trace seams vertex-by-vertex within local neighborhoods. Extensive evaluations show that MeshTailor produces more coherent and structurally regular seam layouts compared to recent optimization-based and learning-based baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2603_27309
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle MeshTailor: Cutting Seams via Generative Mesh Traversal
Ma, Xueqi
Yan, Xingguang
Zhang, Congyue
Huang, Hui
Graphics
Computer Vision and Pattern Recognition
We present MeshTailor, the first mesh-native generative framework for synthesizing edge-aligned seams on 3D surfaces. Unlike prior optimization-based or extrinsic learning-based methods, MeshTailor operates directly on the mesh graph, eliminating projection artifacts and fragile snapping heuristics. We introduce ChainingSeams, a hierarchical serialization of the seam graph that orders chains from global structural cuts down to local details in a coarse-to-fine manner, and a dual-stream encoder that fuses topological and geometric context. Leveraging this hierarchical representation and dual-stream vertex embeddings, our MeshTailor Transformer utilizes an autoregressive pointer layer to trace seams vertex-by-vertex within local neighborhoods. Extensive evaluations show that MeshTailor produces more coherent and structurally regular seam layouts compared to recent optimization-based and learning-based baselines.
title MeshTailor: Cutting Seams via Generative Mesh Traversal
topic Graphics
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2603.27309