Enhancing Raga Identification in Indian Classical Music with FCN-based Models
Shubbh Mewada, Fagun Patel, Sheshang Degadwala, Dhairya Vyas
Abstract
In this study, we provide a novel method for improving raga identification in the field of Indian classical music by making use of Fully Connected Convolutional Networks (FCNs). The broad variety of ragas in this musical tradition are distinguished by subtle melodic and structural differences, making raga identification a difficult task. This study has proposed a novel FCN-based approach, which utilizes deep learning to provide more reliable and effective answers to the problem. The model shows a high degree of accuracy in recognizing and classifying ragas, which bodes well for the potential of automating this difficult activity. This study has performed an extensive review, which included a thorough comparison to prior methods, to prove that our methodology was superior. Our FCN-based model has been shown to perform better than its competitors in this evaluation, demonstrating its potential for real-world use. The rich cultural history of Indian classical music depends on such measures being taken to preserve and disseminate it.