The Music Technology Group (MTG) of the Pompeu Fabra University (UPF) is an internationally recognized research group with 30 years of experience. The group is part of the Department of Information and Communications Technologies, and its research is especially active in topics such as audio signal processing, musical information retrieval, musical interfaces, and computational musicology. The group has extensive experience in research projects both nationally and internationally, and actively works in collaboration with industry. Some technology transfer success stories include Vocaloid, a singing voice synthesiser developed with Yamaha which gained great popularity around the world thanks to the virtual singer Hatsune Miku, and the commercial exploitation of the interactive instrument Reactable, developed at the MTG and used by many popular bands such as Bjork or Coldplay.
Read More
The Music Technology Group (MTG) of the Pompeu Fabra University (UPF) is an internationally recognized research group with 30 years of experience. The group is part of the Department of Information and Communications Technologies, and its research is especially active in topics such as audio signal processing, musical information retrieval, musical interfaces, and computational musicology. The group has extensive experience in research projects both nationally and internationally, and actively works in collaboration with industry. Some technology transfer success stories include Vocaloid, a singing voice synthesiser developed with Yamaha which gained great popularity around the world thanks to the virtual singer Hatsune Miku, and the commercial exploitation of the interactive instrument Reactable, developed at the MTG and used by many popular bands such as Bjork or Coldplay.
In the last years the research team has been involved in some projects related to Artificial intelligence (AI) and its impact in the processes of creating, disseminating, learning and listening to music. The MTG has recently launched a Chair on AI & Music focused on the ethical and social implications of AI in the music sector.
AI has been heralded as a transformative force within the music sector, promising unparalleled opportunities to amplify creativity, accessibility, and efficiency. However, amidst this promise, concerns have arisen from most of the established stakeholders regarding the risks it poses, particularly for artists, prompting calls for robust public regulations. This has triggered an unprecedented public debate in which ethical concerns are taking center stage, underscoring the need for creating AI technologies founded on strong ethical principles.
We should make sure that AI technologies can assist all the music sector stakeholders on their diverse tasks, while placing artists/musicians at the center. Large AI models should aim to capture the essence of music understanding and they should be able to solve specific problems by fine-tuning them. These large AI models should be trained on huge amounts of diverse multimodal music data and their outputs should capture the complex relationships that make up music. The fine-tuned models should support specific tasks related to the creation, production, distribution, access, analysis, or enjoyment of music.