Moving Target Tracking with Physical Feature

Shuai Liu, IMU

Abstract. Recently, since computer vision becomes an important research domain, multimedia understanding and visual tracking of moving target are highlights today. Here, we discuss a novel tracking method for moving target with physical features. Physical features are considered to have strong robustness in real application， and a novel visual tracking method is presented with these physical features. This method transforms these physical features and applies them into real tracking method. Experiments show effectiveness and robustness of this novel method.

 

Efficient Algorithms for Integer Least Squares Problems

Jinming Wen, CNRS

Abstract. Integer least squares (ILS) problems arise from many applications including communications, cryptograph, global positioning systems. One of the commonly used algorithms for solving an ILS problem is sphere decoding which usually consists of two phases: reduction and search. The goal of the reduction is to make the search process more efficient. In this talk, first we introduce some typical algorithms for solving ordinary ILS problems. Then we present some algorithms for solving box-constrained ILS problems which have applications in communications. New effective and efficient reduction algorithms and search algorithms will be presented. Numerical simulation results will be given to show our algorithms can be much faster than existing ones.