Image Patch-Matching with Graph-Based Learning in Street Scenes

• Published in: IEEE transactions on image processing 2023-01, Vol.PP, p.1-1
• Main Authors: She, Rui, Kang, Qiyu, Wang, Sijie, Tay, Wee Peng, Guan, Yong Liang, Navarro, Diego Navarro, Hartmannsgruber, Andreas
• Format: Article
• Language: English
• Subjects: Apexes; Edge joints; Estimation; Feature extraction; Graph matching; graph neural network; Graph theory; Image matching; Image patch-matching; information distance maximization; Kullback-Leibler divergence; Learning; Learning systems; Measurement; Patches (structures); Semantics; Task analysis; visual place recognition
• ISSN: 1057-7149; 1941-0042
• DOI: 10.1109/TIP.2023.3281171

Matching landmark patches from a real-time image captured by an on-vehicle camera with landmark patches in an image database plays an important role in various computer perception tasks for autonomous driving. Current methods focus on local matching for regions of interest and do not take into account spatial neighborhood relationships among the image patches, which typically correspond to objects in the environment. In this paper, we construct a spatial graph with the graph vertices corresponding to patches and edges capturing the spatial neighborhood information. We propose a joint feature and metric learning model with graph-based learning. We provide a theoretical basis for the graph-based loss by showing that the information distance between the distributions conditioned on matched and unmatched pairs is maximized under our framework. We evaluate our model using several street-scene datasets and demonstrate that our approach achieves state-of-the-art matching results.