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

Self-supervised learning has been used to leverage unlabelled data, improvingaccuracy and generalisation of speech systems through the training ofrepresentation models. While many recent works have sought to produce effectiverepresentations across a variety of acoustic domains, languages, modalities andeven simultaneous speakers, these studies have all been limited tosingle-channel audio recordings. This paper presents Spatial HuBERT, aself-supervised speech representation model that learns both acoustic andspatial information pertaining to a single speaker in a potentially noisyenvironment by using multi-channel audio inputs. Spatial HuBERT learnsrepresentations that outperform state-of-the-art single-channel speechrepresentations on a variety of spatial downstream tasks, particularly inreverberant and noisy environments. We also demonstrate the utility of therepresentations learned by Spatial HuBERT on a speech localisation downstreamtask. Along with this paper, we publicly release a new dataset of 100 000simulated first-order ambisonics room impulse responses.

Spatial HuBERT: Self-supervised Spatial Speech Representation Learning for a Single Talker from Multi-channel Audio