Home / AI Smart Systems / ‘AI-Raga’ System
The main objective of ‘AI-RagaGen’ is to show that AI can be used as a tool to generate existing Ragas of Indian Music as well as new Ragas can be generated for Indian music. The system is capable of generating more than 10,000 valid Ragas for Indian music. In the following image it can be seen that in the dialog, the user has selected ‘Ga-Dha’ swaras as the Vadi-Samvadi pair for generating Ragas. It can be seen that the category of the desired Ragas is selected as ‘Oudava-Oudava’ Ragas. With this much information the AIRagaGen system generates all possible Ragas for the given Vadi-Samvadi pair for the particular category of Ragas.
Figure 7a:Raga Generation Prompt
Figure 7b:- Details of Raga Generation Prompt
In the next image the output of the system is displayed. It says ‘Oudava-Oudava’ Ragas for the given Vadi-Samvadi pair are as follows. The first Raga in the generated description happens to be Raga Bhupa. The system provides the detailed description of Raga Bhupa in the form of lists of Anya swaras as a list of ‘rkt’ swaras. It means there are no Anya swaras for this Raga. The next Anu-Anuvadi swara list has only one swara as ‘re’. Re is the only Anu-Anuvadi for the Raga Bhupa. The next list is for Anuvadi swaras which has swara ‘pa’ as primary Anuvadi and swara ‘sa’ as secondary Anuvadi. The next list of Samvadi swara showing ‘dha’ swara and the fifth list has ‘ga’ as the Vadi swara. It also gives the Arohi and Avrohi lists of Varjya swaras that show ‘ma’ and ‘kni’ swaras. Finally, it also provides the Raga Code Name as ‘Gandhapara Susa’ which is a system generated Raga name for Raga Bhupa. In the same manner the description of other possible Ragas is generated. For instance, it can be seen in the image that the next Raga description is given where there is only one change. It has ‘kre’ as the Anu-Anuvadi swara. Other swaras are the same. Accordingly at the end the system generated code name for that Ragas is ‘Gandhapaya Susa’
Figure 8a:Generated Raga Data
The above images provide the description about generating regular standard Ragas as well as all possible valid Ragas. The system also generates ‘Prakirna’ Ragas which somehow do not fit in the regular classification of Ragas. The following image shows the dialog for Prakirna Ragas. A list shows the types of Prakirna Ragas which can be generated by the system as ‘Different Varjya swara Ragas’, ‘Multi-dual swara Ragas’, ‘Tri-Gandhar Ragas’, ‘Chatuswari Ragas’, and other all possible Shadja Gana Samkirna Ragas. The system can also generate all the possible seed Ragas which are also called ‘Beeja Ragas’.
Figure 9:- The various types of Prakirna Ragas that can be generated.
AI-RagaGen is a very useful system that generates all possible valid Ragas based on the GTIM theory. In this sense AI-RagaGen system is the implementation of GTIM theory.