Automatic Language Identification from Audio Signals using LSTM-RNN

The objective of this study is to develop an efficient Language Identification (LID) system using Long Short-Term Memory Recurrent Neural Networks applied to audio signals. Two experiments were conducted to validate the proposed approach. The experimental results demonstrated exceptional performance, with an accuracy of 98% and 97.6 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">%</sup> on the test sets of the first and second experiments, respectively. The models were trained and tested using audio recordings in English, Russian, Turkish, Kyrgyz, and Kazakh languages. These findings suggest that the proposed LID system is highly effective and can be used in various real-world applications.

Перевод пока недоступен