arXiv cs.AI· Aarya Bhave, Anushka Ajay Bhave, Akanksha Narula, Marius Sebastian Shokoui, Alina Hafner·· 2 小时前AI 评分35
2D Gum-Net:基于对称相关与谱池化的无监督稠密弹性图像配准
2D Gum-Net: Unsupervised Dense Elastic Image Registration via Symmetric Correlation and Spectral Pooling
AI 导读
研究者提出 2D Gum-Net,一种完全无监督的端到端稠密弹性形变场预测架构,将原本用于冷冻电子断层扫描 3D 副断层对齐的 Gum-Net 迁移到 2D 弹性配准,用 Dense Spatial Transformer 回归稀疏控制点网格,并以 DCT 谱池化实现约 50 倍 GPU 降采样加速。
正文 · 原文
Abstract:Dense non-rigid registration, the recovery of a pixel-wise deformation field that warps one image into geometric agreement with another, is fundamental across medical imaging, structural biology, and correspondence estimation, yet learning-based approaches face a severe bottleneck: ground-truth deformation fields are notoriously difficult to obtain for real data. We present 2D Gum-Net, an end-to-end architecture that predicts dense elastic deformation fields fully unsupervised. We transfer the Geometric Unsupervised Matching Network (Gum-Net), originally developed for 3D subtomogram alignment in cryo-electron tomography, to 2D elastic registration: all three modules are redesigned in 2D, and the rigid transformation head is replaced by a Dense Spatial Transformer that regresses a sparse control-point grid, lifted to a full-resolution field $\Phi$ by bicubic interpolation. Downsampling uses DCT spectral pooling, implemented as frozen-basis matrix multiplications ($\approx 50\times$ GPU speedup), preserving the fine, quasi-periodic structure that max and average pooling destroy. We evaluate on contact-based fingerprints, a challenging testbed for this task: locally self-similar ridge patterns make correspondence maximally ambiguous, skin pressed against a sensor generates authentic large-displacement elastic distortion, and minutiae act as precise landmarks through which an independent matcher externally verifies the topological correctness of predicted fields. Trained purely on synthetic impressions and evaluated zero-shot on real benchmarks, our method improves structural correlation by up to 15.7% on FVC2004 and raises genuine match scores under an independent matcher while leaving imposter scores essentially unchanged. This suggests that the predicted deformations restore true structure without fabricating correspondences.
| Comments: | 19 pages, 11 figures, 3 tables. Accepted at the NewInML Workshop at the 40th Conference on Neural Information Processing Systems (NeurIPS 2026) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| ACM classes: | I.4.3; I.2.6; I.5.4 |
| Cite as: | arXiv:2611.12805 [cs.CV] |
| (or arXiv:2611.12805v3 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.12805 arXiv-issued DOI via DataCite |
Submission history
From: Aarya Bhave [view email]
[v1]
Mon, 9 Nov 2026 18:11:45 UTC (10,694 KB)
[v2]
Wed, 18 Nov 2026 08:55:23 UTC (9,132 KB)
[v3]
Thu, 19 Nov 2026 10:26:57 UTC (9,132 KB)
来源:arXiv cs.AI · arxiv.org