Improving Deep Learning-based Respiratory Sound Analysis with Frequency Selection and Attention Mechanism

Open paper
Benchmark Model Rank Results
audio-classification-on-icbhiAge-specific CNN-TSA +FBS(BS) 50%freq–ICBHI Score: 62.68Sensitivity: 40.62Specificity: 84.74
audio-classification-on-icbhiAge-specific CNN-TSA +FBS(IS) 50%freq–ICBHI Score: 62.19Sensitivity: 44.93Specificity: 79.44
audio-classification-on-icbhiCNN-TSA +FBS(BS) 50%freq–ICBHI Score: 58.24Sensitivity: 37.48Specificity: 79.00
audio-classification-on-icbhiCNN-TSA +FBS(BS) 75%freq–ICBHI Score: 58.07Sensitivity: 38Specificity: 78.14
audio-classification-on-icbhiCNN-TSA +FBS(IS) 75%freq–ICBHI Score: 60.19Sensitivity: 40Specificity: 80.37

Rank counts only results with a code link.