Deep learning for multiple object tracking: a survey

• Published in: IET computer vision Vol. 13; no. 4; pp. 355 - 368
• Main Authors: Xu, Yingkun, Zhou, Xiaolong, Chen, Shengyong, Li, Fenfen
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 01.06.2019; Wiley
• Subjects: deep learning theory; deep network structures; description enhancement; end-to-end deep network construction; learning (artificial intelligence); multiobject tracking methods; multiple object tracking; network modelling; neural nets; object detection; object tracking; public benchmark test; representation learning; Review Article; tracking conditions; deep network structures; representation learning; public benchmark test; object detection; multiobject tracking methods; multiple object tracking; deep learning theory; description enhancement; end-to-end deep network construction; object tracking; learning (artificial intelligence); neural nets; network modelling; tracking conditions

Deep learning has been proved effective in multiple object tracking, which confronts the difficulties of frequent occlusions, confusing appearance, in-and-out objects, and lack of enough labelled data. Recently, deep learning based multi-object tracking methods make a rapid progress from representat...