VisAdj is a new framework for topology-aware adjacency prediction that introduces an attention-sparse neighbor sampler to adaptively select a high-recall set of candidate node pairs and performs joint edge inference using a line-graph transformer that treats candidate edges as tokens and explicitly models dependencies among incident edges.
Abstract
Learning adjacency matrices from node-link images is a fundamental problem for recovering structured graph information from visual observations. Existing methods typically rely on fixed KNN-based heuristics for candidate edge selection and fail to capture dependencies among edges. To overcome these limitations, we propose VisAdj, a new framework for topology-aware adjacency prediction. VisAdj introduces an attention-sparse neighbor sampler to adaptively select a high-recall set of candidate node pairs and performs joint edge inference using a line-graph transformer that treats candidate edges as tokens and explicitly models dependencies among incident edges. Extensive experiments on synthetic graphs, road networks, and vessel images demonstrate that VisAdj consistently outperforms existing baselines by clear margins.
D2GSL constructs a semantic similarity channel and a spectral feature channel to model node relationships from both local semantic and global spectral views and introduces a hyperadjacency matrix that explicitly models inter-layer node correspondences and enables joint structural reconstruction across channels.
Jun-Chen Zhang, Xuhao Wei, Xiaolei Gu et al.· Computer Modeling in Enginee...· 0 citations
A novel interpretation of sfGOT through the lens of the Hilbert-Schmidt Independence Criterion is provided, showing that minimizing the sfGOT distance between two graphs is equivalent to maximizing statistical dependence between the spectral embedding of their nodes.
This work introduces FALCON (Filtration-based hypergrAph aLignment via Cross-scale Optimal traNsport), an unsupervised optimal-transport framework for hypergraph alignment that constructs a filtration-induced sequence of clique-based co-occurrence dissimilarity matrices and jointly aligns all levels through one shared...
Lutz Oettershagen, Honglian Wang, A. Gionis· 0 citations
Recent graph-based change detection (CD) methods either encode relationships implicitly or rely on post hoc matching, limiting the explicit modeling of object-level spatial and topological relationships under geometric distortions. To address this issue, we propose a CD framework based on a rasterized graph feature map...
Experiments show that the proposed SCoGL variants improve graph recovery and enhance downstream tasks such as graph signal denoising when signal observations are scarce.
Ming-Xiao Liu, Bahar Oveisgharan, Bing-Yan Zou et al.· 0 citations
The Visual-Spatial Latent Graph Network (VSLG-Net), a parameter-compact transformer-based framework with dual-branch for local and global context perception in attention mechanisms, is proposed, which achieves competitive performance on the outdoor YFCC100M benchmark and remains competitive on the indoor SUN3D benchmar...
Wei Lv, Han-Lin Guo, Zhi Shen et al.· Signal, Image and Video Proc...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
