Disentangling Dual-Encoder Masked Autoencoder for Respiratory Sound Classification
By: Peidong Wei, Shiyu Miao, Lin Li
Potential Business Impact:
Helps doctors hear sickness in breathing sounds.
Deep neural networks have been applied to audio spectrograms for respiratory sound classification, but it remains challenging to achieve satisfactory performance due to the scarcity of available data. Moreover, domain mismatch may be introduced into the trained models as a result of the respiratory sound samples being collected from various electronic stethoscopes, patient demographics, and recording environments. To tackle this issue, we proposed a modified MaskedAutoencoder(MAE) model, named Disentangling Dual-Encoder MAE (DDE-MAE) for respiratory sound classification. Two independent encoders were designed to capture disease-related and disease-irrelevant information separately, achieving feature disentanglement to reduce the domain mismatch. Our method achieves a competitive performance on the ICBHI dataset.
Similar Papers
Unmasking Interstitial Lung Diseases: Leveraging Masked Autoencoders for Diagnosis
Image and Video Processing
Helps doctors find lung sickness from scans.
Can Masked Autoencoders Also Listen to Birds?
Machine Learning (CS)
Teaches computers to identify bird songs better.
PSG-MAE: Robust Multitask Sleep Event Monitoring using Multichannel PSG Reconstruction and Inter-channel Contrastive Learning
Machine Learning (CS)
Helps doctors better understand sleep and breathing.