LSTM, RNN and GAN Models Effectiveness for Flute Music Generation: A Thorough Expedition

Within the realm of algorithmic music composition, machine learning-powered systems alleviate the necessity of painstakingly crafting composition rules by hand. This paper embarks on a symphonic journey into the realm of automatic music generation, where the magic of deep learning orchestrates musical masterpieces. The resulting compositions take shape as enchanting sequences of ABC notes. In our modern age, the harmony of music technology harmoniously resonates with the grandeur of vast datasets. When it comes to crafting musical wonders through the wonders of deep learning, maestros often choose the eloquent rhythms LSTM or the graceful melodies of Vanilla GAN and Cycle-GAN models for their symphonic canvas. In this melodious endeavour, where music becomes a sequence of moments, different GAN models shines as the virtuoso of choice, playing the sweetest of notes. Also the results of evaluation by human experts shows the ability of different GAN models suggested in this article. The results mentioned in the article shows the generated music is more compare to human generated music and is human audible.