This paper presents two techniques for improving human body tracking within the particle filtering scheme. Both techniques explore the use of auxiliary measurements. The first technique uses optical flow cues to improve the sampling distribution. The second technique involves the detection of individual body parts, namely the hand, head and torso; and using these detection results to provide ad...

Faces usually are the most interesting objects in certain categories of video like home videos and news clips. In this paper a novel sensor fusion based face tracking system is presented that tracks faces in compressed video, and aids automatic video indexing. Tracking is done by fusing the measurements from three independent sensors – motion and colour based trackers (derived from [2]) and a f...

Fast estimation and tracking of the principal subspace of a sequence of random vectors is a classic problem, widely encountered in areas such as radar, sonar and speech processing, data compression, data filtering, parameter estimation, pattern recognition, neural analysis, wireless communications, to name just a few. Among the most robust algorithms for subspace tracking there are the so calle...

In this paper, an original method named GNG-T, extended from GNG-U algorithm by [1] is presented. The method performs continuously vector quantization over a distribution that changes over time. It deals with both sudden changes and continuous ones, and is thus suited for video tracking framework, where continuous tracking is required as well as fast adaptation to incoming and outgoing people. ...

We consider nonlinear filtering applications to target tracking based on a vector of multi-scaled models where some of the processes are rapidly mean reverting to their local equilibria. We focus attention on target tracking problems because multiple scaled models with fast mean-reversion (FMR) are a simple way to model latency in the response of tracking systems. The main results of this paper...

Object tracking is very important for automatic indexing of video content. This work shows tracking of objects directly using compressed MPEG video data. Two sensors, one using motion vectors and the other using DCT coefficients obtained from compressed video stream, provide measurements for the location of the object being tracked. The optimal estimate from the two measurements is found using ...

The paper presents a cascade control scheme for Permanent Magnet Synchronous Motors (PMSMs) based on a novel reduced order state observer which provides accurate speed tracking performance. The asymptotic vanishing of both the observation error and the speed tracking error is proven. The observer design allows a residual signal to be defined, able to detect faults using Park’s vector currents. ...

This paper analyzes the performance of an optimal bounding ellipsoid (OBE) algorithm for tracking time-varying parameters with incrementally bounded time variations. A linear state-space model is used, with the time-varying parameters represented by the state vector. The OBE algorithm exhibits a selective update property for the time and observation-update equations, and necessary and suf-cient...