Automated lung sound analysis and respiratory disease classification using deep learning. Features a multi-branch audio pipeline extracting Mel Spectrogram, Constant-Q Transform (CQT), and Continuous Wavelet Transform (CWT) features for spatio-temporal classification on the ICBHI 2017 dataset.
machine-learning deep-learning signal-processing pytorch audio-classification wavelet-transform cnn-lstm constant-q-transform feature-fusion biomedical-signal-processing respiratory-disease audio-preprocessing icbhi-2017 lung-sound-analysis
-
Updated
Jun 22, 2026 - Python