Emotion based music style change using deep learning

A system for changing music of a video game based on a player's emotion, comprising:

• a processor;
• and
• a memory comprising instructions stored thereon, which when executed by the processor, causes the processor to perform:
• receiving indicators of emotion comprising in-game attributes of a player in a video game;
• receiving original music from the video game;
• determining an original tone of the original music using a model trained to recognize the original tone from the original music;
• and
• generating transformed music from the original music by a variational autoencoder, wherein the variational autoencoder:
• compresses the original music into a dimensional representation and reconstructs the dimensional representation based at least in part on the indicators of emotion.

A non-transitory computer-readable storage medium comprising instructions stored thereon, which when executed by the one or more processors, cause the one or more processors to perform operations for changing music of a video game based on a player's emotion, the operations comprising:

• receiving indicators of emotion comprising in-game attributes of a player in a video game;
• receiving original music from the video game;
• determining an original tone of the original music using a model trained to recognize the original tone from the original music;
• transforming the original tone of the original music to a transformed tone by a variational autoencoder;
• and
• generating transformed music from the original music based on the transformed tone by the variational autoencoder, wherein the variational autoencoder:
• compresses the original music into a dimensional representation and reconstructs the dimensional representation based at least in part on the indicators of emotion.