Learning the decision function for speaker verification

Explores the possibility of replacing the usual thresholding decision rule of log likelihood ratios used in speaker verification systems by more complex and discriminant decision functions based for instance on linear regression models or support vector machines. Current speaker verification systems, based on generative models such as HMMs or Gaussian mixture models, can indeed easily be adapted to use such decision functions. Experiments on both text dependent and text independent tasks always yielded performance improvements and sometimes significantly.