吉林大学学报（理学版）2026，Vol.64Issue(2)：284-290,7.DOI:10.13413/j.cnki.jdxblxb.2024503
融合欧氏与双曲几何的深度度量学习方法
Deep Metric Learning Method Combining Euclidean and Hyperbolic Geometry
摘要
Abstract
Aiming at the isotropy problem caused by the widespread use of cosine metrics in proxy-based deep metric learning methods,we proposed a deep metric learning method that integrated Euclidean geometry and hyperbolic geometry.By introducing hyperbolic geometry with advantages in hierarchical modeling,a local hyperbolic loss function was designed in hyperbolic space,and the distribution prior of hyperbolic space was used to initialize proxy points reasonably.During training process,the local neighborhood proxy points corresponding to each sample were dynamically optimized,thereby effectively enhancing the inter-class discriminative ability of the model in local regions.Experimental results show that the proposed method exhibits significant performance improvements on multiple standard image retrieval datasets,thus validating the effectiveness of blending different geometric properties for enhancing discriminative performance in metric learning.关键词
深度度量学习/双曲几何/图像检索/计算机视觉Key words
deep metric learning/hyperbolic geometry/image retrieval/computer vision分类
信息技术与安全科学引用本文复制引用
张书达,李慧盈..融合欧氏与双曲几何的深度度量学习方法[J].吉林大学学报（理学版）,2026,64(2):284-290,7.基金项目
吉林省科技发展计划项目(批准号:20230201089GX). （批准号:20230201089GX）
