Convolutional Auto‑Encoder for Variable Length Respiratory Sound Compression and Reconstruction

Shuailin Tao, Jinhai Hu, Wang Ling Goh, Yuan Gao

2024 IEEE Biomedical Circuits and Systems Conference (BioCAS 2024), 2024

Autoencoders · Variable-Length Sequences

Overview

An autoencoder-based approach to compressing and reconstructing variable-length audio sequences, with a composite loss designed to preserve reconstruction quality. The method was evaluated on respiratory sound recordings in the BioCAS 2024 Grand Challenge.