BobLiu20/YOLOv3_PyTorch

About the loss computation

Opened this issue · 1 comments

chrisway613 commented

I was confused about why use BCE to compute the loss of 'x' & 'y', could it be MSE?

YOLOv3_PyTorch/nets/yolo_loss.py

Line 55 in c6b4837

loss_x = self.bce_loss(x * mask, tx * mask)

BigDataHa commented

I was confused about why use BCE to compute the loss of 'x' & 'y', could it be MSE?

YOLOv3_PyTorch/nets/yolo_loss.py

Line 55 in c6b4837

loss_x = self.bce_loss(x * mask, tx * mask)

I have same question with you