الفهرس Only 14 pages are availabe for public view
 |  | from | 89 |  |  |
| | | from | 89 | | |
Abstract
Stuttering represents the main speech disfluency problem affects on the smooth- ness of a person speech with the most two common stuttering disfluencies events are repetitions and prolongations. In Speech Pathology field , the speech pathologists(SP) are required to keep track of people suffer stuttering starting from analyzing each one case in terms of measuring the rate of stuttering, identifying types of stuttering events the stutterer suffer from and other factors that considered to be complementary in- formation which helps for deciding the best treatment plan. It is most desired for SPs to classify these disfluencies automatically rather than manually classification, which is a subjective, time-consuming task, and depends also on speech language pathologists experience. This automatic classi- fication of the events will help pathologists to go ahead with the treatment plan easily through measuring its success at each treatment stage which turns finally into a low rate of stuttering.Several classification techniques has been used before for the automatic clas- sification of stuttering events with a good classification accuracies. However, it is still a point of discussion in that field.The aim of the current work was to introduce a new usage for the I-Vector approach, the well known in speaker recognition/verification applications. where it is used here for speech disfluencies automatic classification independently on speaker in a way to test its performance in such speech applications and trying to enhance the classification accuracies which will sound in enhancing the speech applications (speech recognition/ speaker recognition) related to the stuttering problem