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 Tools annual

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 monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Noise suppression (NS) models have been widely applied to enhance speechquality. Recently, Deep Learning-Based NS, which we denote as Deep NoiseSuppression (DNS), became the mainstream NS method due to its excellingperformance over traditional ones. However, DNS models face 2 major challengesfor supporting the real-world applications. First, high-performing DNS modelsare usually large in size, causing deployment difficulties. Second, DNS modelsrequire extensive training data, including noisy audios as inputs and cleanaudios as labels. It is often difficult to obtain clean labels for training DNSmodels. We propose the use of knowledge distillation (KD) to resolve bothchallenges. Our study serves 2 main purposes. To begin with, we are among thefirst to comprehensively investigate mainstream KD techniques on DNS models toresolve the two challenges. Furthermore, we propose a novelAttention-Based-Compression KD method that outperforms all investigatedmainstream KD frameworks on DNS task.

ABC-KD: Attention-Based-Compression Knowledge Distillation for Deep Learning-Based Noise Suppression