PERANCANGAN SISTEM KEAMANAN LINGKUNGAN PENGENAL SUARA KULKUL DENGAN MENGGUNAKAN METODE DEEP LEARNING

Abstract

Advances in information technology provide benefits for people's lives today. Before the existence of information technology, people still use traditional communication media called kulkul. Kulkul is a communication tool that is used by hitting. Over time, people began to leave this culture because of the many other information systems used as communication media. However, nowadays people in their teens still do not know what the sound of the kulkul means. This is due to the absence of a kulkul voice recognition website. The method of data collection in this study was the observation method of a kulkul worker located in  Denjalan Subvillage, Batubulan, Sukawati, Gianyar. To build a kulkul voice recognition system, the author uses a deep learning method. In this system there are 2 processes, namely training and classification. The training is used for the system to learn to recognize the sound of the kulkul and the classification to determine the category of the kulkul sound. Based on the classification carried out, the results obtained is testing accuracy 85%.