سال انتشار: ۱۳۹۰

محل انتشار: پنجمین کنفرانس بین المللی پیشرفتهای علوم و تکنولوژی

تعداد صفحات: ۵

نویسنده(ها):

Navid Samimi Behbahan – Sama Technical and Vocational Training College, Islamic Azad University, Behbahan

چکیده:

In this paper, we are going to investigate the topic of recognition of certain words in a noisy environment. This topic is very important due to loss of the driver’s concentration while working with the audio system of a car. The words which was mentioned by the driver are: next (next track), were prior (prior track), increase (volume) and decrease (volume). To exact the feature of each frame, we recruited Mel Frequency Cepstral Coefficient, First derivative Second derivative with audio energy (in aggregate 39 features). The Hidden Morkov Model algorithm was recruits signify that the proposal algorithm set has a high level performance