Speaker recognition and speaker segmentation : application over the Internet