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

Inspired by the recent progress in self-supervised learning for computervision, in this paper we introduce DeLoRes, a new general-purpose audiorepresentation learning approach. Our main objective is to make our networklearn representations in a resource-constrained setting (both data andcompute), that can generalize well across a diverse set of downstream tasks.Inspired from the Barlow Twins objective function, we propose to learnembeddings that are invariant to distortions of an input audio sample, whilemaking sure that they contain non-redundant information about the sample. Toachieve this, we measure the cross-correlation matrix between the outputs oftwo identical networks fed with distorted versions of an audio segment sampledfrom an audio file and make it as close to the identity matrix as possible. Weuse a combination of a small subset of the large-scale AudioSet dataset andFSD50K for self-supervised learning and are able to learn with less than halfthe parameters compared to state-of-the-art algorithms. For evaluation, wetransfer these learned representations to 9 downstream classification tasks,including speech, music, and animal sounds, and show competitive results underdifferent evaluation setups. In addition to being simple and intuitive, ourpre-training algorithm is amenable to compute through its inherent nature ofconstruction and does not require careful implementation details to avoidtrivial or degenerate solutions. Furthermore, we conduct ablation studies onour results and make all our code and pre-trained models publicly availablehttps://github.com/Speech-Lab-IITM/DeLoRes.

DeLoRes: Decorrelating Latent Spaces for Low-Resource Audio Representation Learning