A Deep Learning Architecture with Spatio-Temporal Focusing for Detecting Respiratory Anomalies

Dat Ngo (Autor:in und Vortragende:r), Lam Pham, Huy Phan, Minh Tran, Delaram Jarchi

Publikation: Beitrag in Buch oder TagungsbandVortrag mit Beitrag in TagungsbandBegutachtung

Abstract

This paper presents a deep learning system applied for detecting anomalies from respiratory sound recordings. Our system initially performs audio feature extraction using Continuous Wavelet transformation. This transformation converts the respiratory sound input into a two-dimensional spectrogram where both spectral and temporal features are presented. Then, our proposed deep learning architecture inspired by the Inception-residual-based backbone performs the spatio-temporal-focusing and multi-head attention mechanism to classify respiratory anomalies. In this work, we evaluate our proposed models on the benchmark SPRSound (The Open-Source SJTU Paediatric Respiratory Sound) database proposed by the IEEE BioCAS 2023 challenge. As regards the Score computed by an average between the average score and harmonic score, our robust system has achieved Top-1 performance with Scores of 0.810, 0.667, 0.744, and 0.608 in Tasks 1-1, 1-2, 2-1, and 2-2, respectively.
OriginalspracheEnglisch
TitelArtificial Intelligence BioMedical Circuits And Systems For Health
Seiten1-5
ISBN (elektronisch)979-8-3503-0026-0
DOIs
PublikationsstatusVeröffentlicht - 2023
Veranstaltung2023 IEEE Biomedical Circuits and Systems Conference (BioCAS) - Toronto, ON, Toronto, Kanada
Dauer: 19 Okt. 202321 Okt. 2023

Konferenz

Konferenz2023 IEEE Biomedical Circuits and Systems Conference (BioCAS)
Land/GebietKanada
StadtToronto
Zeitraum19/10/2321/10/23

Research Field

  • Multimodal Analytics

Fingerprint

Untersuchen Sie die Forschungsthemen von „A Deep Learning Architecture with Spatio-Temporal Focusing for Detecting Respiratory Anomalies“. Zusammen bilden sie einen einzigartigen Fingerprint.

Diese Publikation zitieren