Automatic Recognition of Dialects of Himachal Pradesh Using MFCC &GMM
Ayushi Dogra, Amit Kaul, RavinderNath Sharma · 2019
In the present globalized society, people from all over the world are coming and working together. The biggest hurdle in this process is the medium of interaction, as all individuals do not speak a common language. A way to circumvent this problem is to develop a machine based system for recognition of different languages and subsequently employing automatic machine translation mechanism for communication between people from diverse language or dialect groups. A MFCC and GMM based approach for recognition of dialects of Himachal Pradesh has been presented in this paper. Around seven dialects spoken in different regions of Himachal Pradesh have been used in this study. The results obtained shows that dialects spoken in adjacent regions have significant overlap and are difficult to distinguish. Accuracy for offline and online tests was respectively around 80% and 70%.