Abstract:
In this study a method for artificially composing visual accompaniment for
music pieces is proposed. We analyze whether the proposed method composes visual
accompaniments that are comparable in quality to visual accompaniments made by a
human artist. It was found that visual accompaniments composed by the proposed
methods are judged significantly lower in quality than their human-made
counterparts. Additionally, it was found that the performance of the proposed method
did not differ significantly from a pseudo-random approach to composing visual
accompaniments. Despite these results, this method might provide a framework for
future research on this topic.