Oktoechos Classification and Generation of Liturgical Music using Deep Learning Frameworks

Abstract

In the oktoec̄hos tradition, liturgical hymns are sung in eight modes or eight colours (known as eight’niram’ in Indian liturgy). In this paper, recurrent neural network (RNN) models are used for oktoec̄hos genre classification with the help of musical texture features (MTF) and i-vectors. The performance of the proposed approaches is evaluated using a newly created corpus of liturgical music in the South Indian language-Malayalam. Long short-term memory (LSTM)-based and gated recurrent unit (GRU)-based experiments report an average classification accuracy of 83.76% and 77.77%, respectively, with a significant margin over the i-vector-DNN framework. The experiments demonstrate the potential of RNN models in learning temporal information through MTF in recognising eight modes of oktoec̄hos system. Furthermore, since the Greek liturgy and Gregorian chant also share similar musical traits with Syrian tradition, the musicological insights observed can potentially be applied to those traditions. The generation of oktoec̄hos genre music style is discussed using an encoder-decoder framework. The quality of the generated files is evaluated using a perception test.