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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Polyphonic events are the main error source of audio event detection (AED)systems. In deep-learning context, the most common approach to deal with eventoverlaps is to treat the AED task as a multi-label classification problem. Bydoing this, we inherently consider multiple one-vs.-rest classificationproblems, which are jointly solved by a single (i.e. shared) network. In thiswork, to better handle polyphonic mixtures, we propose to frame the task as amulti-class classification problem by considering each possible labelcombination as one class. To circumvent the large number of arising classes dueto combinatorial explosion, we divide the event categories into multiple groupsand construct a multi-task problem in a divide-and-conquer fashion, where eachof the tasks is a multi-class classification problem. A network architecture isthen devised for multi-class multi-task modelling. The network is composed of abackbone subnet and multiple task-specific subnets. The task-specific subnetsare designed to learn time-frequency and channel attention masks to extractfeatures for the task at hand from the common feature maps learned by thebackbone. Experiments on the TUT-SED-Synthetic-2016 with high degree of eventoverlap show that the proposed approach results in more favorable performancethan the common multi-label approach.

Polyphonic audio event detection: multi-label or multi-class multi-task classification problem?