Enhancement of ELDA Tracker Based on CNN Features and Adaptive Model Update

Changxin Gao1, Huizhang Shi1, Jin-Gang Yu2, Nong Sang1

1National Key Laboratory of Science and Technology on Multispectral Information Processing,

School of Automation, Huazhong University of Science and Technology, Wuhan, 430074, China

2Department of Computer Science and Engineering,

University of Nebraska-Lincoln, Lincoln, NE, 68503, USA

Abstract

Appearance representation and the observation model are the most important components in designing a robust visual tracking algorithm for video-based sensors. Additionally, the exemplar-based linear discriminant analysis (ELDA) model has shown good performance in object tracking. Based on that, we improve the ELDA tracking algorithm by deep convolutional neural network (CNN) features and adaptive model update. Deep CNN features have been successfully used in various computer vision tasks. Extracting CNN features on all of the candidate windows is time consuming. To address this problem, a two-step CNN feature extraction method is proposed by separately computing convolutional layers and fully-connected layers. Due to the strong discriminative ability of CNN features and the exemplar-based model, we update both object and background models to improve their adaptivity and to deal with the tradeoff between discriminative ability and adaptivity. An object updating method is proposed to select the “good” models (detectors), which are quite discriminative and uncorrelated to other selected models. Meanwhile, we build the background model as a Gaussian mixture model (GMM) to adapt to complex scenes, which is initialized offline and updated online. The proposed tracker is evaluated on a benchmark dataset of 50 video sequences with various challenges. It achieves the best overall performance among the compared state-of-the-art trackers, which demonstrates the effectiveness and robustness of our tracking algorithm.

Related papers

  • Changxin Gao, Huizhang Shi, Jin-Gang Yu*, Nong Sang. "Enhancement of ELDA Tracker Based on CNN Features and Adaptive Model Update," Sensors, 16(4):545, 2016 (paper)
  • Changxin Gao, Feifei Chen, Jin-Gang Yu, Rui Huang, Nong Sang. "Exemplar-based Linear Discriminant Analysis for Robust Object Tracking," ICIP, 2014, pp.388-392. (paper) (bibtex)
  • Changxin Gao, Feifei Chen, Jin-Gang Yu, Rui Huang, Nong Sang*. "Robust Visual Tracking Using Exemplar-based Detectors," IEEE T CSVT, to apear, 2016 (paper) (bibtex) (project)
  • DOI:10.1109/TCSVT.2015.2513700

Source Code

Matlab code(without object model update and background model update). (including off-line background model)

Results

Video results (12 videos).

Plots figures on CVPR2013 benchmark.

Result locations on CVPR2013 benchmark: ELDA_CNN (without object model update and background model update).