Implementasi Speech Recognition pada Aplikasi Virtual Reality Training untuk Meningkatkan Kemampuan Public Speaking Mahasiswa Teknologi Informasi

Abstract

This study aims to develop a Virtual Reality Training application by implementing speech recognition to improve the public speaking skills of information technology students. The system was developed using Unity and the Wav2Vec2 model for Bahasa Indonesia from Hugging Face in analyzing user verbal responses. This study was conducted in three training sessions, designed to measure the development of users' public speaking skills gradually. Data from 17 participants were collected and analyzed using the Friedman test to measure significant differences between sessions and continued with the Wilcoxon Signed-Rank test to compare the three sessions. The results of the analysis showed that there were significant differences in the aspects of words per minute, volume, speaking duration, eye contact, and hand movements. This proves that the application of speech recognition using virtual reality not only helps improve the public speaking skills of information technology students, but also shows that repeated practice contributes to improved performance.