Skip to main content
. 2022 Nov 29;10(12):2396. doi: 10.3390/healthcare10122396
Algorithm 2 Symptoms detection procedures
Input: Only face image
Output: The symptom recognition
  •   1:

    Using CSPDarknet53 transforms feature maps of different sizes in different convolutional layers.

  •   2:

    Use SPP to transform feature maps of any size into feature vectors of fixed size to improve the perceptual field of view.

  •   3:

    Use PAN to blend three feature maps of different sizes.

  •   4:

    Pass the result of step 4 to the head of YOLO and output the three feature maps: (19,19,num_anchor*(num_classes + 5)), (38,38,num_anchor*(num_classes + 5)), (76,76,num_anchor*(num_classes + 5)).

  •   5:

    Three different-size feature maps are used to calculate the prediction BB.

  •   6:

    The IoU is compared with the ground-truth BB to calculate the IoU loss.

  •   7:

    End.