Adaptive Game Soundtrack Generation Based on Music Transcription

Abstract

This research aims to bring new prospects for adaptive game soundtrack generation, which theoretically can increase the involvement of players into the gaming process. To investigate the problem of adaptive soundtrack generation, human-centered evaluation has to be done. Furthermore, an important part of this research is a collaborative soundtrack generation with the participation of composers and artificial intelligence, which is able to decrease expenses of game developing companies. For soundtrack customization, this work suggests a joint generation of music by human and artificial intelligence and its further intuitive modification during the game based on in-game actions.