I'm trying to train the model myself. In the paper, the loss is defined as l1 loss plus a gradient loss in “Learning the depths of moving people by watching frozen people”. But I found that they calculate the gradient loss in log space. Do you use log depth and gt as well? Besides, for the l1 loss, |(Dxs −Dgt)|, are the output and gt both in the range [0, 255]?
Hi,
About the log space no, I did not use the log space nor the mask (thank you for pointing out this, I’m going to make it clear). For the second question yes, but it should work even with a different range
Hi,
I'm trying to train the model myself. In the paper, the loss is defined as l1 loss plus a gradient loss in “Learning the depths of moving people by watching frozen people”. But I found that they calculate the gradient loss in log space. Do you use log depth and gt as well? Besides, for the l1 loss, |(Dxs −Dgt)|, are the output and gt both in the range [0, 255]?
Thanks.