CLNet: A Compact Latent Network for Fast Adjusting Siamese Trackers

Document Type

Conference Proceeding

Publication Title

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

Abstract

In this paper, we provide a deep analysis for Siamese-based trackers and find that the one core reason for their failure on challenging cases can be attributed to the problem of decisive samples missing during offline training. Furthermore, we notice that the samples given in the first frame can be viewed as the decisive samples for the sequence since they contain rich sequence-specific information. To make full use of these sequence-specific samples, we propose a compact latent network to quickly adjust the tracking model to adapt to new scenes. A statistic-based compact latent feature is proposed to efficiently capture the sequence-specific information for the fast adjustment. In addition, we design a new training approach based on a diverse sample mining strategy to further improve the discrimination ability of our compact latent network. To evaluate the effectiveness of our method, we apply it to adjust a recent state-of-the-art tracker, SiamRPN++. Extensive experimental results on five recent benchmarks demonstrate that the adjusted tracker achieves promising improvement in terms of tracking accuracy, with almost the same speed. The code and models are available at https://github.com/xingpingdong/CLNet-tracking.

First Page

378

Last Page

395

DOI

10.1007/978-3-030-58565-5_23

Publication Date

11-12-2020

Keywords

Fast adjustment, Latent feature, Sample mining, Sequence-specific, Siamese tracker

Comments

IR conditions: non-described

Share

COinS