HOW CAN DEEP LEARNING MODELS IDENTIFY AND COMPARE MELODIC SIMILARITIES BETWEEN SONGS TO DETECT POTENTIAL COMPOSITIONAL COPYRIGHT INFRINGEMENT?

The music industry and its evolution are closely tied to the growth of technology, including both the creation and distribution of music in the 21st century. However, such technology has yet to impact the protection of music IP. Judicial courts handle the cases of compositional copyright infringement in a subjective manner–taking in the opinions of many factfinders as if the two songs “sound” similar with regard to explicit tune, style, and authorship, to arrive at a final verdict. Thus, this research explores the extent to which deep-learning models can bring forth objectivity in the proceedings of court hearings to provide a holistic view of the entire case, therefore, increasing the certainty of a court’s verdict to be justifiable.