Song Title Generation

Thesis Type Master
Thesis Status
Student Simon Hachmeier
Number of Students
Thesis Supervisor
Research Field

More and more research efforts are being made towards automatic music composition and generation. We can already foresee that this will be an important topic for both research and industry. However, what is still missing, are methods to automatically generate song titles that match the generated music. 

To take a first step towards this goal, in this thesis, we aim to treat the song title generation task as a text summarization problem. I.e., we use a dataset of song lyrics and aim to generate the title tracks based on deep neural network architectures. In a next step, we aim to add MIDI data such that the model not only considers lyrics but also audio content information for the generation process.