Possible bug in region loss calculation

[sidml]

Hi

I was going through the region_loss code. I found a possible bug in build_targets function,

for t in range(nTrueBox):
    if target[b][t][1] == 0:
        break

I think it should have been if target[b][t][0] == 0 because the class is supposed to be n index 0. Index 1 contains the x coordinate. In the current code, the loss will be updated even if its a background class. Please correct me if my interpretation is incorrect. Thanks.