Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 2
Encontrada(s) 1 página(s)

Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico antes de ser publicado, versión del autor (Preprint)

A diffusion-inspired training strategy for singing voice extraction in the waveform domain

  • Plaja-Roglans, Genís
  • Miron, Marius
  • Serra, Xavier
This work has been accepted at the 23rd International Society for Music Information Retrieval Conference (ISMIR 2022), at Bengaluru, India. December 4-8, 2022., Notable progress in music source separation has been achieved using multi-branch networks that operate on both temporal and spectral domains. However, such networks tend to be complex and heavy-weighted. In this work, we tackle the task of singing voice extraction from polyphonic music signals in an end-to-end manner using an approach inspired by the training procedure of denoising diffusion models. We perform unconditional signal modelling to gradually convert an input mixture signal to the corresponding singing voice or accompaniment. We use fewer parameters than the state-of-the-art models while operating on the waveform domain, bypassing phase-related problems. More concisely, we train a non-causal WaveNet using a diffusion-inspired strategy improving the said network for singing voice extraction and obtaining performance comparable to the end-to-end state-of-the-art on MUSDB18. We further report results on a non-MUSDB-overlapping version of MedleyDB and the multi-track audio of the Saraga Carnatic dataset showing good generalization, and run perceptual tests of our approach. Code, models, and audio examples are made available., This work was carried out under the projects Musical AI - PID2019-111403GB-I00/AEI/10.13039/501100011033 and NextCore - RTC2019-007248-7 funded by the Spanish Ministerio de Ciencia, Innovación y Universidades (MCIU) and the Agencia Estatal de Investigación (AEI).
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico antes de ser publicado, versión del autor (Preprint)

BAF: an audio fingerprinting dataset for broadcast monitoring

  • Cortès, Guillem
  • Ciurana, Alex
  • Molina, Emilio
  • Miron, Marius
  • Meyers, Owen
  • Six, Joren
  • Serra, Xavier
This work has been accepted at the 23rd International Society for Music Information Retrieval Conference (ISMIR 2022), at Bengaluru, India. December 4-8, 2022., Audio Fingerprinting (AFP) is a well-studied problem in music information retrieval for various use-cases e.g. content-based copy detection, DJ-set monitoring, and music excerpt identification. However, AFP for continuous broadcast monitoring (e.g. for TV & Radio), where music is often in the background, has not received much attention despite its importance to the music industry. In this paper (1) we present BAF, the first public dataset for music monitoring in broadcast. It contains 74 hours of production music from Epidemic Sound and 57 hours of TV audio recordings. Furthermore, BAF provides cross-annotations with exact matching timestamps between Epidemic tracks and TV recordings. Approximately, 80% of the total annotated time is background music. (2) We benchmark BAF with public state-of-the-art AFP systems, together with our proposed baseline PeakFP: a simple, non-scalable AFP algorithm based on spectral peak matching. In this benchmark, none of the algorithms obtain a F1-score above 47%, pointing out that further research is needed to reach the AFP performance levels in other studied use cases. The dataset, baseline, and benchmark framework are open and available for research., This research is part of NextCore – New generation of music monitoring technology (RTC2019-007248-7), funded by the Spanish Ministerio de Ciencia e Innovación and the Agencia Estatal de Investigación. Also, has received support from Industrial Doctorates plan of the Secretaria d’universitats i Recerca, Departament d’Empresa i Coneixement de la Generalitat de Catalunya, grant agreement No. DI46-2020.