Metraipolis › archive › dc_d29ef9e1caa8
2D Gum-Net: Unsupervised Dense Elastic Image Registration via Symmetric Correlation and Spectral Pooling — arXiv:2611.12805
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.12805
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-10T18:44:18.294Z
- SHA-256 of the archived text
eed4f2d44ce3776728884b5859786b370bb66329247ebb30e4723fe4fac9ef39- Stable URL of this page
https://metraipolis.com/documento/dc_d29ef9e1caa8
How to cite
2D Gum-Net: Unsupervised Dense Elastic Image Registration via Symmetric Correlation and Spectral Pooling — arXiv:2611.12805 — arXiv (en), https://arxiv.org/abs/2611.12805 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d29ef9e1caa8 (Metraipolis archive dc_d29ef9e1caa8, sha256 eed4f2d44ce3776728884b5859786b370bb66329247ebb30e4723fe4fac9ef39)
BibTeX
@misc{dcd29ef9e1caa8,
title = {2D Gum-Net: Unsupervised Dense Elastic Image Registration via Symmetric Correlation and Spectral Pooling — arXiv:2611.12805},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.12805},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d29ef9e1caa8; sha256 eed4f2d44ce3776728884b5859786b370bb66329247ebb30e4723fe4fac9ef39},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.12805. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
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. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2611.12805 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-10T18:37:37.707Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).