Sound recognition has become necessary in many fields, such as security, music, and advertising. However, sound recognition remains challenging due to the complexity of sounds and the variety of backgrounds.

Music monitoring (audio fingerprinting)  is used to identify songs based on the unique characteristics of the audio recording. Although it has been studied for some time, it is a technological challenge when considering noisy environments such as concerts, bars, and outdoor shows or environments where music plays in the background, as is often the case on radio and television.

The PhD of Guillem Cortès addresses solving this challenge. It is carried out under the supervision of Prof. Xavier Serra (MTG) and Dr. Emilio Molina (BMAT). It is the second Industrial PhD between BMAT and the MTG, after Blai Meléndez. The project aims to investigate how to improve music monitoring with deep learning algorithms and make them more robust in multiple scenarios, as well as create tools to encourage research in this field.

Read the interview here:

https://doctoratsindustrials.gencat.cat/la-musica-i-la-tecnologia-es-fusionen-en-un-projecte-entre-bmat-i-la-upf/