Search

Clustering aims to group n data samples into k clusters. In this paper, we reformulate the clustering problem into an integer optimization problem and propose a recurrent neural network with n × k neurons to solve it. We prove the stability and convergence of the proposed recurrent neural network theoretically. Moreover, clustering experiments demonstrate that the proposed clustering algorithm based on the recurrent neural network can achieve the better clustering performance than existing clustering algorithms. [ABSTRACT FROM AUTHOR] more...

Robust and accurate visual tracking is challenging as targets undergo significant changes in appearance by scale variance, occlusion and fast motion. We propose a novel tracking framework, called scalable spatiotemporal visual tracking algorithm (SSVT). First, we construct the Direction Prediction Model (DPM) to predict the spatiotemporal correlation of the target in the next frame. That will efficiently narrow down the search area and improve the accuracy of spatial location. Then, Occlusion Detection algorithm (ODA) is presented to overcome the wrong updates stemming from the region of interest (ROI) based on the estimated direction and Kalman filter. Finally, the multi-scale pyramid kernelized correlation filter (MSPKCF) is presented in tracking to realize the adaptive adjustment of the varying scales of the targets and the ROI size. Extensive experiments on OTB100 and VOT2016 datasets demonstrate that our tracker performs favorably against state-of-the-art trackers, which can effectively reduce computation redundancy and improve tracking accuracy. [ABSTRACT FROM AUTHOR] more...