In music monitoring systems, fingerprinting technology is used in an audio sample to identify similar items in a database that includes: songs, melodies, sound effects, etc. The acoustic digital fingerprint serves to monitor the use of certain musical works and performances on radio programmes, records, CDs, etc. This identification is used, for example, in complying with licensing and copyright.

Despite the commercial success of fingerprinting technologies in music monitoring systems, they still have key technological limitations that prevent the correct detection of music in many special contexts, such as live music, when music is in the background or for low quality recordings, and even when fingerprinting technology itself is not applicable and cover identification technology must be used.

A new project: NextCore Next generation of music monitoring technology aims to research and develop qualitative improvements to musical fingerprinting identification technologies and audio coverage of BMAT, a spinoff of the Music Technology Group (MTG) at UPF, founded in 2005.  The project is funded by the Spanish Ministry of Science and Innovation and coordinated by the BMAT. It involved the MTG, which is led by Xavier Serra at the Department of Information and Communication Technologies (DTIC) at UPF.

“Restrictions in digital fingerprinting technology in monitoring applied to music mean that a significant percentage of the music played is not reported correctly, which generates significant financial losses for the authors and the music industry and are the basis of current technological competition in the field of digital rights management”, says Xavier Serra, principal investigator of the project.

Thus, the main objective of NextCore will be to research and develop qualitative improvements in digital fingerprinting identification technology and audio coverage of BMAT in three main areas:

• identifying recorded music when it undergoes substantial alterations (e.g., reinterpreted by a DJ) or when the signal is captured in unfavourable environments (e.g., via a microphone with background noise),
• identifying musical works at unpublished performances (e.g., live events, recorded versions of songs),
• Detecting the context in which the performance is broadcast (e.g., radio or TV aggregators).

Within NextCore, the MTG will work on learning architecture to improve music detection pre-processing audio signals captured in situations where music is in the background or when the recording is of poor quality, and research new architectures for the scalable detection of musical coverage.

BMAT will be the business partner of the NextCore project. Founded in 2005 as a spinoff from the MTG, it specializes in monitoring music worldwide through television, radio, digital sites and services. BMAT provides music monitoring and reporting services to more than 100 collective management organizations (CMO) and over 2,000 record labels, publishers, digital service providers (DSPs) and broadcasters in 134 countries, becoming a neutral source of data available to any stakeholder involved in the music value cycle.

The Music Technology Group (MTG) at Pompeu Fabra University in Barcelona, of the Department of Information and Communication Technologies, conducts research on such issues as audio signal processing, music information retrieval, music interfaces, and computational musicology.

Reference project: 

NEXTCORE: Next generation of music monitoring technology

Project funded by the Spanish Ministry of Science and Innovation (Reference: RTC2019-007248-7)

Duration: 1 October 2020 to 31 September 2023

Partners: BMAT (Coordinator), MTG-UPF

PI of the MTG: Xavier Serra