Saved in:
Bibliographic Details
Main Authors: Zhu, Haokai, Qu, Bo, Cao, Si-Yuan, Zhang, Runmin, Chen, Shujie, Yang, Bailin, Shen, Hui-Liang
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2509.07662
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908527708078080
author Zhu, Haokai
Qu, Bo
Cao, Si-Yuan
Zhang, Runmin
Chen, Shujie
Yang, Bailin
Shen, Hui-Liang
author_facet Zhu, Haokai
Qu, Bo
Cao, Si-Yuan
Zhang, Runmin
Chen, Shujie
Yang, Bailin
Shen, Hui-Liang
contents Previous deep image registration methods that employ single homography, multi-grid homography, or thin-plate spline often struggle with real scenes containing depth disparities due to their inherent limitations. To address this, we propose an Exponential-Decay Free-Form Deformation Network (EDFFDNet), which employs free-form deformation with an exponential-decay basis function. This design achieves higher efficiency and performs well in scenes with depth disparities, benefiting from its inherent locality. We also introduce an Adaptive Sparse Motion Aggregator (ASMA), which replaces the MLP motion aggregator used in previous methods. By transforming dense interactions into sparse ones, ASMA reduces parameters and improves accuracy. Additionally, we propose a progressive correlation refinement strategy that leverages global-local correlation patterns for coarse-to-fine motion estimation, further enhancing efficiency and accuracy. Experiments demonstrate that EDFFDNet reduces parameters, memory, and total runtime by 70.5%, 32.6%, and 33.7%, respectively, while achieving a 0.5 dB PSNR gain over the state-of-the-art method. With an additional local refinement stage,EDFFDNet-2 further improves PSNR by 1.06 dB while maintaining lower computational costs. Our method also demonstrates strong generalization ability across datasets, outperforming previous deep learning methods.
format Preprint
id arxiv_https___arxiv_org_abs_2509_07662
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle EDFFDNet: Towards Accurate and Efficient Unsupervised Multi-Grid Image Registration
Zhu, Haokai
Qu, Bo
Cao, Si-Yuan
Zhang, Runmin
Chen, Shujie
Yang, Bailin
Shen, Hui-Liang
Computer Vision and Pattern Recognition
Previous deep image registration methods that employ single homography, multi-grid homography, or thin-plate spline often struggle with real scenes containing depth disparities due to their inherent limitations. To address this, we propose an Exponential-Decay Free-Form Deformation Network (EDFFDNet), which employs free-form deformation with an exponential-decay basis function. This design achieves higher efficiency and performs well in scenes with depth disparities, benefiting from its inherent locality. We also introduce an Adaptive Sparse Motion Aggregator (ASMA), which replaces the MLP motion aggregator used in previous methods. By transforming dense interactions into sparse ones, ASMA reduces parameters and improves accuracy. Additionally, we propose a progressive correlation refinement strategy that leverages global-local correlation patterns for coarse-to-fine motion estimation, further enhancing efficiency and accuracy. Experiments demonstrate that EDFFDNet reduces parameters, memory, and total runtime by 70.5%, 32.6%, and 33.7%, respectively, while achieving a 0.5 dB PSNR gain over the state-of-the-art method. With an additional local refinement stage,EDFFDNet-2 further improves PSNR by 1.06 dB while maintaining lower computational costs. Our method also demonstrates strong generalization ability across datasets, outperforming previous deep learning methods.
title EDFFDNet: Towards Accurate and Efficient Unsupervised Multi-Grid Image Registration
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2509.07662