Abstract:Convolutional neural networks (CNN) based tracking approaches have shown favorable performance in recent benchmarks. Nonetheless, the chosen CNN features are always pre-trained in different task and individual components in tracking systems are learned separately, thus the achieved tracking performance may be suboptimal. Besides, most of these trackers are not designed towards real-time applications because of their time-consuming feature extraction and complex optimization <a class="link-external link-http" href="http://details.In" rel="external noopener nofollow">this http URL</a> this paper, we propose an end-to-end framework to learn the convolutional features and perform the tracking process simultaneously, namely, a unified convolutional tracker (UCT). Specifically, The UCT treats feature extractor and tracking process both as convolution operation and trains them jointly, enabling learned CNN features are tightly coupled to tracking process. In online tracking, an efficient updating method is proposed by introducing peak-versus-noise ratio (PNR) criterion, and scale changes are handled efficiently by incorporating a scale branch into network. The proposed approach results in superior tracking performance, while maintaining real-time speed. The standard UCT and UCT-Lite can track generic objects at 41 FPS and 154 FPS without further optimization, respectively. Experiments are performed on four challenging benchmark tracking datasets: OTB2013, OTB2015, VOT2014 and VOT2015, and our method achieves state-of-the-art results on these benchmarks compared with other real-time trackers.

Real-Time Online Tracking Via a Convolution-Based Complementary Model

Continuity-Discrimination Convolutional Neural Network for Visual Object Tracking

Real-time visual tracking based on an appearance model and a motion mode

Online Object Tracking Based on CNN with Spatial-Temporal Saliency Guided Sampling

An Ensemble of Complementary Models for Deep Tracking

Robust Visual Tracking Via Multiple Discriminative Models with Object Proposals

Online Scale Adaptive Visual Tracking Based on Multilayer Convolutional Features

Robust and Real-Time Visual Tracking Based on Complementary Learners.

Online Video Tracking Using Collaborative Convolutional Networks

Online Tracking With Convolutional Neural Networks

UCT: Learning Unified Convolutional Networks for Real-time Visual Tracking

A Complementary Tracking Model with Multiple Features.

Learning a Real-Time Generic Tracker Using Convolutional Neural Networks

Online Parallel Framework for Real-Time Visual Tracking.

Residual Attention Convolutional Network for Online Visual Tracking

Online Adaptive Complementation Tracker

Exploiting Weak Mask Representation with Convolutional Neural Networks for Accurate Object Tracking.

High Performance Visual Object Tracking with Unified Convolutional Networks

Multiple Convolutional Neural Networks for Online Visual Tracking.

Visual tracking utilizing robust complementary learner and adaptive refiner.

Real-Time Tracking Based on Keypoints and Discriminative Correlation Filters