No abstract

In a multi-target multi-measurement environment, knowledge of the measuresnent-Lo-trk assignments is typally unavailable to the traking algorithm. In this r, a strictly probabiIist apoah to the measiwesnag-so-trk assignment problem is taken. Measurements are it assigned to tracks as in traditional multi-hypothesis tracking (Mill) algoriihms instead, the prObability that eh measurement belongs to eh track is estimated using a maximum Iikeiihond (ML) algorithm derived by the method of Expectafion-Maximization (EM). These measurement-to-trk probability estimates arc intrinsic to the multi-target tracker called the probabilistic multi-hypothesis trking (PMWF) algorithm. Unlike MHT algorithms, the PMHT algorithm does zt maintain explicit hypothesis lists. The PMHT algorithm is computatkxially practical because it requires neither enumaation of measurement-to-trk assignments nor pruning.

A maximum likelihood method is presented for training probabilistic neural networks (PNN's) using a Gaussian kernel, or Parzen window. The proposed training algorithm enables general nonlinear discrimination and is a generalization of Fisher's method for linear discrimination. Important features of maximum likelihood training for PNN's are: 1) it economizes the well known Parzen window estimator while preserving feedforward NN architecture, 2) it utilizes class pooling to generalize classes represented by small training sets, 3) it gives smooth discriminant boundaries that often are "piece-wise flat" for statistical robustness, 4) it is very fast computationally compared to backpropagation, and 5) it is numerically stable. The effectiveness of the proposed maximum likelihood training algorithm is assessed using nonparametric statistical methods to define tolerance intervals on PNN classification performance.

No abstract

No abstract

No abstract

No abstract

No abstract