Multi-object tracking with discriminant correlation filter based deep learning tracker

Article type: Research Article

Authors: Yang, Tao | Cappelle, Cindy | Ruichek, Yassine^* | El Bagdouri, Mohammed

Affiliations: CIAD, University Bourgogne Franche-Comté, UTBM, 90010 Belfort, France

Correspondence: [*] Corresponding author: Yassine Ruichek, CIAD, University Bourgogne Franche-Comté, UTBM, 90010 Belfort, France. E-mail: yassine.ruichek@utbm.fr.

Abstract: In this paper, we extend the discriminant correlation filter (DCF) based deep learning tracker to multi-object tracking. For each object, we use an individual tracker to estimate the position. Two different pre-trained networks are used as feature extractors, respectively. The response peak and oscillation are both considered to validate the tracking. When the object is lost, the discriminative appearance model achieved by DCF is considered as a part of the feature representation between the object and detection for data association. In order to validate our method, we analyze and test our approach on the MOT2D2015 and MOT17 benchmarks for multiple pedestrian tracking. The results show that our approach performs superiorly against several recent state-of-the-art online multi-object trackers.

Keywords: Multi-object tracking, correlation filter, convolutional neural network, data association

DOI: 10.3233/ICA-180596

Journal: Integrated Computer-Aided Engineering, vol. 26, no. 3, pp. 273-284, 2019