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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

This paper presents a novel approach for indoor acoustic source localizationusing microphone arrays and based on a Convolutional Neural Network (CNN). Theproposed solution is, to the best of our knowledge, the first published work inwhich the CNN is designed to directly estimate the three dimensional positionof an acoustic source, using the raw audio signal as the input informationavoiding the use of hand crafted audio features. Given the limited amount ofavailable localization data, we propose in this paper a training strategy basedon two steps. We first train our network using semi-synthetic data, generatedfrom close talk speech recordings, and where we simulate the time delays anddistortion suffered in the signal that propagates from the source to the arrayof microphones. We then fine tune this network using a small amount of realdata. Our experimental results show that this strategy is able to producenetworks that significantly improve existing localization methods based on\textit{SRP-PHAT} strategies. In addition, our experiments show that our CNNmethod exhibits better resistance against varying gender of the speaker anddifferent window sizes compared with the other methods.

Towards End-to-End Acoustic Localization using Deep Learning: from Audio Signal to Source Position Coordinates