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

The ongoing biodiversity crisis, driven by factors such as land-use changeand global warming, emphasizes the need for effective ecological monitoringmethods. Acoustic monitoring of biodiversity has emerged as an importantmonitoring tool. Detecting human voices in soundscape monitoring projects isuseful both for analysing human disturbance and for privacy filtering. Despitesignificant strides in deep learning in recent years, the deployment of largeneural networks on compact devices poses challenges due to memory and latencyconstraints. Our approach focuses on leveraging knowledge distillationtechniques to design efficient, lightweight student models for speech detectionin bioacoustics. In particular, we employed the MobileNetV3-Small-Pi model tocreate compact yet effective student architectures to compare against thelarger EcoVAD teacher model, a well-regarded voice detection architecture ineco-acoustic monitoring. The comparative analysis included examining variousconfigurations of the MobileNetV3-Small-Pi derived student models to identifyoptimal performance. Additionally, a thorough evaluation of differentdistillation techniques was conducted to ascertain the most effective methodfor model selection. Our findings revealed that the distilled models exhibitedcomparable performance to the EcoVAD teacher model, indicating a promisingapproach to overcoming computational barriers for real-time ecologicalmonitoring.

Efficient speech detection in environmental audio using acoustic recognition and knowledge distillation