Robust tracking-by-detection using a selection and completion mechanism

作     者：Ruochen Fan Fang-Lue Zhang Min Zhang Ralph R.Martin 

作者机构：Tsinghua University School of Engineering and Computer Science Victoria University of Wellington Center of Mathematical Sciences and Applications Harvard University School of Computer Science and Informatics Cardiff University 

出 版 物：《Computational Visual Media》 (计算可视媒体（英文版）)

年 卷 期：2017年第3卷第3期

页      面：285-294页

核心收录：

学科分类：08[工学] 080203[工学-机械设计及理论] 0835[工学-软件工程] 0802[工学-机械工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by the National Natural Science Foundation of China (Project No. 61521002) the General Financial Grant from the China Postdoctoral Science Foundation (Grant No. 2015M580100) a Research Grant of Beijing Higher Institution Engineering Research Center, and an EPSRC Travel Grant 

主　　题：object tracking detection proposal selection trajectory completion 

摘      要：It is challenging to track a target continuously in videos with long-term occlusion,or objects which leave then re-enter a *** tracking algorithms combined with onlinetrained object detectors perform unreliably in complex conditions, and can only provide discontinuous trajectories with jumps in position when the object is occluded. This paper proposes a novel framework of tracking-by-detection using selection and completion to solve the abovementioned problems. It has two components, tracking and trajectory completion. An offline-trained object detector can localize objects in the same category as the object being tracked. The object detector is based on a highly accurate deep learning model. The object selector determines which object should be used to re-initialize a traditional tracker. As the object selector is trained online,it allows the framework to be adaptable. During completion, a predictive non-linear autoregressive neural network completes any discontinuous *** tracking component is an online real-time algorithm, and the completion part is an after-theevent mechanism. Quantitative experiments show a significant improvement in robustness over prior stateof-the-art methods.