Skip to content

AsthmaSCELNet: A Lightweight Supervised Contrastive Embedding Learning Framework For Asthma Classification Using Lung Sounds (INTERSPEECH 2024)

Notifications You must be signed in to change notification settings

rsarka34/AsthmaSCELNet-INTERSPEECH

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

70 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AsthmaSCELNet

AsthmaSCELNet: A Lightweight Supervised Contrastive Embedding Learning Framework For Asthma Classification Using Lung Sounds

Authors: Arka Roy, Udit Satija

Paper Link Paper Link YouTube Link

Abstract

Asthma is one of the most prevalent respiratory disorders, which can be identified by different modalities such as speech, wheezing of lung sounds (LSs), spirometric measures, etc. In this paper, we propose AsthmaSCELNet, a lightweight supervised contrastive embedding learning framework, to classify asthmatic LSs by providing adequate classification margin across the embeddings of healthy and asthma LS, in contrast to vanilla supervised learning. Our proposed framework consists of three steps: pre-processing, melspectrogram extraction, and classification. The AsthmaSCELNet consists of two stages: embedding learning using a lightweight embedding extraction backbone module that extracts compact embedding from the melspectrogram, and classification by the learnt embeddings using multi-layer perceptrons. The proposed framework achieves an accuracy, sensitivity, and specificity of 98.54%, 98.27%, and 98.73% respectively, that outperforms existing methods based on LSs and other modalities.

Methodology

20

Dataset

Dataset Link: Paper Link image

Results

image

Performance

image

Cite as:

A. Roy, U. Satija, "AsthmaSCELNet: A Lightweight Supervised Contrastive Embedding Learning Framework for Asthma Classification Using Lung Sounds", in Proc. INTERSPEECH 2023, 5431-5435, doi: 10.21437/Interspeech.2023-428.

@inproceedings{roy23_interspeech,
  author={Arka Roy and Udit Satija},
  title={{AsthmaSCELNet: A Lightweight Supervised Contrastive Embedding Learning Framework for Asthma Classification Using Lung Sounds}},
  year=2023,
  booktitle={Proc. INTERSPEECH 2023},
  pages={5431--5435},
  doi={10.21437/Interspeech.2023-428}
}

Releases

No releases published

Packages

No packages published

Languages