Speech Analysis of Chhattisgarhi Dialects using Wavelet Transformation and Mel Frequency Cepstral Coefficient

The native dialects of distinct regions of Chhattisgarh are known as that state's dialects. In the year 2000, Chhattisgarhi was adopted as the state's official language and became a part of India. Using wavelet analysis techniques, this paper compared and identified speech patterns in diverse Chhattisgarhi dialects, which are suitable for challenging speech signal analysis. The dataset was obtained from a typical Chhattisgarh native, and this work presents the dialects used in different regions of the state. Wavelet transformations are used to extract parameters from a speech corpus, after which relations are built. The experiment was carried out with the aid of a Jupiter notepad and the Librosa library. Excellent features for audio signal analysis are offered by Librosa. Mel-frequency cepstral coefficient is the most effective method for digitally altering the audio stream and are utilised for feature extraction. This work extracted Wavelet Transform and Mel-frequency Cpstral co-efficients both type of features.