An adaptive music generation architecture for games based on the deep learning Transformer mode
This addresses the need for personalized and dynamic music generation in video games, but it is incremental as it builds on existing Transformer and layering techniques.
The paper tackles the problem of generating adaptive music for video games by proposing an architecture based on the Transformer model, which customizes music to player preferences and game situations using an arousal-valence emotion model, though no concrete performance numbers are provided.
This paper presents an architecture for generating music for video games based on the Transformer deep learning model. Our motivation is to be able to customize the generation according to the taste of the player, who can select a corpus of training examples, corresponding to his preferred musical style. The system generates various musical layers, following the standard layering strategy currently used by composers designing video game music. To adapt the music generated to the game play and to the player(s) situation, we are using an arousal-valence model of emotions, in order to control the selection of musical layers. We discuss current limitations and prospects for the future, such as collaborative and interactive control of the musical components.