Audio-based Assessment in Determining Language

Abstract

The most apparent way humans communicate is verbal with the use of Language. With language being so complex, the average person would not be able to recognize another language besides their native own. But one thing is universal for all languages spoken throughout the world and it is that they produce sound. The average human being can hear sounds with frequencies ranging from 20 Hz to 20,000 Hz and with an intensity above the standard threshold of hearing. Hence, speaking produces sound with frequencies in between the said range. This paper aims to determine the language spoken of a given voice signal. By analyzing the given voice signal, its frequency range can be determined. Languages are spoken differently therefore have different frequency ranges. By using a filter to remove any ambient sounds from the input voice signal, its equivalent spectrogram reading gives an accurate range of its frequencies.