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/

Zendy Plus

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

Zendy Tools

Zendy Open

While multitask and transfer learning has shown to improve the performance ofneural networks in limited data settings, they require pretraining of the modelon large datasets beforehand. In this paper, we focus on improving theperformance of weakly supervised sound event detection in low data and noisysettings simultaneously without requiring any pretraining task. To that extent,we propose a shared encoder architecture with sound event detection as aprimary task and an additional secondary decoder for a self-supervisedauxiliary task. We empirically evaluate the proposed framework for weaklysupervised sound event detection on a remix dataset of the DCASE 2019 task 1acoustic scene data with DCASE 2018 Task 2 sounds event data under 0, 10 and 20dB SNR. To ensure we retain the localisation information of multiple soundevents, we propose a two-step attention pooling mechanism that provides atime-frequency localisation of multiple audio events in the clip. The proposedframework with two-step attention outperforms existing benchmark models by22.3%, 12.8%, 5.9% on 0, 10 and 20 dB SNR respectively. We carry out anablation study to determine the contribution of the auxiliary task and two-stepattention pooling to the SED performance improvement.

Improving weakly supervised sound event detection with self-supervised auxiliary tasks