FURI | Spring 2022

VivyNet: A New Approach to Creating AI-Generated Music

Advances in machine learning (ML) have allowed researchers to simulate human behaviors. One such behavior is content generation. This field of artificial media creation is difficult to perfect as there are multiple ways an ML model can perform poorly. This difficulty affects the quality of the generated content and does not truly mimic the human nature of the outputted source. This research aims to achieve the following: utilize poems as an input to artificially create classical musical compositions, and research different ways to solve this dilemma by utilizing novel techniques and various ML training methods.