English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In spite of the progress in music source separation research, the smallamount of publicly-available clean source data remains a constant limitingfactor for performance. Thus, recent advances in self-supervised learningpresent a largely-unexplored opportunity for improving separation models byleveraging unlabelled music data. In this paper, we propose a self-supervisedlearning framework for music source separation inspired by the HuBERT speechrepresentation model. We first investigate the potential impact of the originalHuBERT model by inserting an adapted version of it into the well-known DemucsV2 time-domain separation model architecture. We then propose atime-frequency-domain self-supervised model, Pac-HuBERT (for primitive auditoryclustering HuBERT), that we later use in combination with a Res-U-Net decoderfor source separation. Pac-HuBERT uses primitive auditory features of music asunsupervised clustering labels to initialize the self-supervised pretrainingprocess using the Free Music Archive (FMA) dataset. The resulting frameworkachieves better source-to-distortion ratio (SDR) performance on the MusDB18test set than the original Demucs V2 and Res-U-Net models. We furtherdemonstrate that it can boost performance with small amounts of superviseddata. Ultimately, our proposed framework is an effective solution to thechallenge of limited clean source data for music source separation.

Pac-HuBERT: Self-Supervised Music Source Separation via Primitive Auditory Clustering and Hidden-Unit BERT