Deep regression tracking with shrinkage loss

Xiankai Lu, Chao Ma, Bingbing Ni, Xiaokang Yang, Ian Reid, Ming Hsuan Yang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

37 Citations (Scopus)

Abstract

Regression trackers directly learn a mapping from regularly dense samples of target objects to soft labels, which are usually generated by a Gaussian function, to estimate target positions. Due to the potential for fast-tracking and easy implementation, regression trackers have recently received increasing attention. However, state-of-the-art deep regression trackers do not perform as well as discriminative correlation filters (DCFs) trackers. We identify the main bottleneck of training regression networks as extreme foreground-background data imbalance. To balance training data, we propose a novel shrinkage loss to penalize the importance of easy training data. Additionally, we apply residual connections to fuse multiple convolutional layers as well as their output response maps. Without bells and whistles, the proposed deep regression tracking method performs favorably against state-of-the-art trackers, especially in comparison with DCFs trackers, on five benchmark datasets including OTB-2013, OTB-2015, Temple-128, UAV-123 and VOT-2016.

Original languageEnglish
Title of host publicationComputer Vision – ECCV 2018 - 15th European Conference, 2018, Proceedings
EditorsVittorio Ferrari, Cristian Sminchisescu, Yair Weiss, Martial Hebert
PublisherSpringer Verlag
Pages369-386
Number of pages18
ISBN (Print)9783030012632
DOIs
Publication statusPublished - 2018
Event15th European Conference on Computer Vision, ECCV 2018 - Munich, Germany
Duration: 2018 Sep 82018 Sep 14

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11218 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other15th European Conference on Computer Vision, ECCV 2018
CountryGermany
CityMunich
Period18/9/818/9/14

Bibliographical note

Funding Information:
Acknowledgments. This work is supported in part by the National Key Research and Development Program of China (2016YFB1001003), NSFC (61527804, 61521062, U1611461, 61502301, and 61671298), the 111 Program (B07022), and STCSM (17511105401 and 18DZ2270700). C. Ma and I. Reid acknowledge the support of the Australian Research Council through the Centre of Excellence for Robotic Vision (CE140100016) and Laureate Fellowship (FL130100102). B. Ni is supported by China’s Thousand Youth Talents Plan. M.-H. Yang is supported by NSF CAREER (1149783).

Publisher Copyright:
© 2018, Springer Nature Switzerland AG.

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Fingerprint Dive into the research topics of 'Deep regression tracking with shrinkage loss'. Together they form a unique fingerprint.

Cite this