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

Existing weakly supervised sound event detection (WSSED) work has notexplored both types of co-occurrences simultaneously, i.e., some sound eventsoften co-occur, and their occurrences are usually accompanied by specificbackground sounds, so they would be inevitably entangled, causingmisclassification and biased localization results with only clip-levelsupervision. To tackle this issue, we first establish a structural causal model(SCM) to reveal that the context is the main cause of co-occurrence confoundersthat mislead the model to learn spurious correlations between frames andclip-level labels. Based on the causal analysis, we propose a causalintervention (CI) method for WSSED to remove the negative impact ofco-occurrence confounders by iteratively accumulating every possible context ofeach class and then re-projecting the contexts to the frame-level features formaking the event boundary clearer. Experiments show that our method effectivelyimproves the performance on multiple datasets and can generalize to variousbaseline models.

Improving Weakly Supervised Sound Event Detection with Causal Intervention