xuebinqin
commented
4 years ago Thank you for your interests. Because the iou was defined as 1-iou in its definition part. The reason we use ssim loss as 1-Ssim is that the ssim code was borrowed from another implementation as shown in the readme. We don’t want to change its original definition so that use 1-ssim.
On Dec 26, 2019, at 10:30 AM, pp00704831 notifications@github.com wrote:
Hello, I want to ask some questions about iou loss and ssim loss. When I saw the code basnet_train.py, I found that :
ssim_out = 1 - ssim_loss(pred,target) iou_out = iou_loss(pred,target)
I think that iou loss still need to be subtracted by 1 like that: iou_out = 1 - iou_loss(pred,target)
But why it do not need to do that? Thanks you!
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.
Hello, I want to ask some questions about iou loss and ssim loss. When I saw the code basnet_train.py, I found that :
ssim_out = 1 - ssim_loss(pred,target) iou_out = iou_loss(pred,target)
I think that iou loss still need to be subtracted by 1 like that: iou_out = 1 - iou_loss(pred,target)
But why it do not need to do that? Thanks you!