| 研究実績の概要 |
We have developed a novel deep learning system for efficient detection and monitoring of cough events in audio recordings. First, we perform voice activity detection to eliminate audio silences and focus on relevant segments. Next, we employ a cough classification technique to identify the presence of cough within those audio segments. Finally, we implement cough event detection using a high performance classification-regression fusion method. Our approach differs from the traditional audio event detection methods in several notable ways: (1) we incorporate a teacher-student framework for the training of our detection model, (2) instead of relying on specific audio features such as MFCC or Mel Spectrogram, our end-to-end system takes the raw audio signal directly as input and outputs the cough boundary timings, (3) the proposed method is general enough to be used for various other sound event monitoring tasks. The evaluation of our detection model in a 3-fold cross-validation experiment demonstrates its strong performance and robustness. Specifically, we achieved cough event detection F1, Recall, and Precision scores of 87.21%, 82.91%, and 92.36% respectively. The comparative cough monitoring assessment of our system against other platforms, including the Leicester Cough Monitor, demonstrates our system's superiority by achieving the lowest average hourly symmetric mean absolute error (sMAPE) of 8.48%.
|
