Adaptive_Fusion_RGBD_Saliency_Detection icon indicating copy to clipboard operation
Adaptive_Fusion_RGBD_Saliency_Detection copied to clipboard

Published 20 hours ago • verified publisher icon Lucia-Ningning

dataset.py

Open Mrlong12 opened this issue 6 years ago • 2 comments

where is the dataset.py I dont konw how to Normalize the depth image.

Mrlong12 avatar Feb 22 '19 02:02 Mrlong12

If you have depth images like those in data/depth/, just feed them into the network. If you have raw depth data, normalize it into [0,1](it should be [0,255] in image format), 1 stands for the closest position while 0 stands for the farthest. The following codes may be helpful to you:

tempD = rawDepth - min(rawDepth(:))
normD = 1 - tempD / max(tempD(:))

Lucia-Ningning avatar Feb 22 '19 04:02 Lucia-Ningning

thank you！when I look at your code , I do not know your total loss how to calculate.how many are the weights of three loss, when you calculate the total loss?

Mrlong12 avatar Feb 25 '19 01:02 Mrlong12