Research Article

Deep Learning versus Professional Healthcare Equipment: A Fine-Grained Breathing Rate Monitoring Model

Table 2

Model parameters. The convolutional layer parameters are denoted as “〈convolutional type〉 conv 〈receptive field size〉 − 〈number of channels〉.” Also, the parameters of recurrent layer are denoted as “〈uni or bi〉 − 〈number of hidden units〉,” where “uni” denoted unidirectional recurrent and “bi” denoted bidirectional recurrent. The ReLU activation function is not shown for brevity.

Model configuration

DeepFilter 1DeepFilter 2DeepFilter 3
6 weight7 weight6 weight
LayersLayersLayers
Input (882  50 sequence)Input (220  4  50 spectrogram)
One-dimension-conv9-16Two-dimension-conv3-16Two-dimension-conv3-16
Two-dimension-conv3-32Two-dimension-conv3-32
Mean pool
FC-512
FC-512
FC-512
Uni-128Uni-128Bi-128
Uni-128Uni-128
Output: sigmod
Loss: cross-entropy