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

This thesis focuses on dealing with the task of acoustic scene classification(ASC), and then applied the techniques developed for ASC to a real-lifeapplication of detecting respiratory disease. To deal with ASC challenges, thisthesis addresses three main factors that directly affect the performance of anASC system. Firstly, this thesis explores input features by making use ofmultiple spectrograms (log-mel, Gamma, and CQT) for low-level featureextraction to tackle the issue of insufficiently discriminative or descriptiveinput features. Next, a novel Encoder network architecture is introduced. TheEncoder firstly transforms each low-level spectrogram into high-levelintermediate features, or embeddings, and thus combines these high-levelfeatures to form a very distinct composite feature. The composite or combinedfeature is then explored in terms of classification performance, with differentDecoders such as Random Forest (RF), Multilayer Perception (MLP), and Mixtureof Experts (MoE). By using this Encoder-Decoder framework, it helps to reducethe computation cost of the reference process in ASC systems which make use ofmultiple spectrogram inputs. Since the proposed techniques applied for generalASC tasks were shown to be highly effective, this inspired an application to aspecific real-life application. This was namely the 2017 Internal Conference onBiomedical Health Informatics (ICBHI) respiratory sound dataset. Building uponthe proposed ASC framework, the ICBHI tasks were tackled with a deep learningframework, and the resulting system shown to be capable at detectingrespiratory anomaly cycles and diseases.

Robust Deep Learning Frameworks for Acoustic Scene and Respiratory Sound Classification