Kernalised Multi-Resolution Convnet for Visual Tracking

Di Wu, Wenbin Zou, Xia Li, Yong Zhao; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2017, pp. 66-73

Discriminative Correlation Filters (DCF) have demonstrated excellent performance for high-speed visual object tracking. Built upon their seminar work, there has been a plethora of recent improvements relying on convolutional neural network. In this paper, we go beyond the conventional DCF framework and propose a Kernalised Multi-resolution Convnet (KMC) formulation that utilises hierarchical response maps to directly output the target movement. The performance of this transfer learning paradigm is on par with a variety of state-of-the-art approaches and when directly deployed the learnt network to predict the unseen challenging UAV tracking dataset. Moreover, the transfered multi-reslution CNN renders it possible to be integrated into the RNN temporal learning framework, therefore opening the door on the end-to-end temporal deep learning (TDL) for visual tracking.