GenMelody Automatic Melody Generation For A Given Phrase

Abstract

Melody song generation for a given phrase is a crucial task in symbolic music composition and computational creativity. This paper presents GenMelody, an automatic melody song generation system for different Indian languages designed to produce coherent and expressive melodies for input text phrases. The study explores various approaches, including rule-based systems, statistical models, and deep learning-based techniques. We analyze the effectiveness of Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, and Transformer-based architectures in melody synthesis. Additionally, we discuss hybrid approaches that integrate music theory constraints with AI-generated compositions. The experimental results highlight the advantages and challenges of different methods, providing insights into future research directions in AI-driven music generation.