Deep metric learning for image retrieval in smart city development. (October 2021)

• Record Type: Journal Article
• Authors: Liu, Qi; Li, Wenhan; Chen, Zhiyuan; Hua, Bin
• Is Part Of: Sustainable cities and society. Volume 73(2021)
• Publication Date: 2021-10
• Subjects: Smart city -- Image retrieval -- Deep metric learning; Sustainable urban development -- Periodicals; Sustainable buildings -- Periodicals; Urban ecology (Sociology) -- Periodicals; 307.76
• DOI: 10.1016/j.scs.2021.103067
• Languages: English
• ISSNs: 2210-6707

Abstract

Graphical abstract: Highlights: Instance proxy captures much more global data distribution information for learning. Label-smoothing technique for positive proxies to preserve intra-class structure. New soft-instance-label loss boosts performance with SOTA results achieved. Abstract: Deep metric learning (DML ) aims to learn a consistent distance embedding where an anchor is closer within the same category than others. It underpins a variety of essential and significant tasks in the development of smart city including face recognition, landmark retrieval, pedestrian detection, person/vehicle re-identification, and so on. Traditional pair-based DML methods try to make full use of the data-to-data relations within a (mini-)batch, but they cannot grasp the data distribution information due to the batch size limitation. On the other hand, proxy-based DML schemes use different proxies to approximate the data distribution. However, the proxies are too sample to represent the intra-category variance. In this paper, we propose a simple but effective method, named soft-instance-label proxy, for embedding learning. It can capture the globe data distribution information while depicting the detailed intra-class data structure. The state-of-the-art empirical results on three public image retrieval benchmarks and two backbone networks demonstrate the superiority of our proposed method. Our Soft-instance-label proxy method can have a Recall @ 1 improvement of 2.4% with Googlenet, largely surpassing the current state-of-art-methods while demonstrating great potential in the development of smart city.