arXiv cs.AI· Geonwoo Baek, Ikbeom Jang·· 1 天前AI 评分20
vSV-ViT：面向阿尔茨海默病皮层表面学习的可变尺寸超顶点视觉 Transformer
vSV-ViT: Variable-size SuperVertex Vision Transformer for Cortical Surface Learning in Alzheimer's Disease
AI 导读
研究者提出可变尺寸超顶点（vSV）划分方法，并据此设计 vSV-ViT——通过填充、掩码感知的 patch 嵌入与顶点位置嵌入处理 vSV 的视觉 Transformer，用于皮层表面学习。在阿尔茨海默病脑影像数据集上，以皮层厚度与皮层曲率网格为输入，vSV-ViT 在 AD 相关分类与性别预测任务上预测准确率优于近期基于表面的模型。代码已公开，论文被 ACCV 2026 接收。
正文 · 原文
Abstract:Learning from 3D meshes is challenging because the data reside on non-Euclidean surfaces embedded in 3D space. This is particularly evident in domains such as brain cortical surface analysis, where existing models typically rely on ROI-agnostic, face-based, or fixed-size patches. Such patches can duplicate boundary vertices, conflate anatomically distinct regions, or incorporate non-cortical vertices such as those of the medial wall. We develop variable-size supervertex (vSV) partitioning, which groups cortical surface vertices into vSV patches similar to superpixels in images. Building on these vSVs, we design Variable-size SuperVertex Vision Transformer (vSV-ViT), a Vision Transformer that processes vSVs through padding, mask-aware patch embedding, and vertex position embedding. We demonstrate the benefits of the proposed framework through a quantitative analysis of vertex assignment in surface partitioning and downstream tasks using Alzheimer's disease brain image datasets. For downstream tasks, we use meshes of cortical thickness and cortical curvature as input to various encoder models and evaluate classification performance. The superior predictive accuracy of vSV-ViT over recent surface-based models supports its effectiveness for cortical surface learning, with AD-related and sex prediction serving as downstream validation. Extending the framework to other cortical surface applications and broader 3D mesh domains remains future work. Code is available at this https URL.
| Comments: | Accepted to ACCV 2026 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.26514 [cs.CV] |
| (or arXiv:2605.26514v3 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.26514 arXiv-issued DOI via DataCite |
Submission history
From: Geonwoo Baek [view email]
[v1]
Tue, 26 May 2026 03:57:33 UTC (13,405 KB)
[v2]
Sun, 4 Oct 2026 11:08:03 UTC (3,620 KB)
[v3]
Wed, 11 Nov 2026 09:07:46 UTC (3,791 KB)
来源：arXiv cs.AI · arxiv.org