Absolute difference instead of Normalized difference?

[meet-minimalist]

For the loc_loss calculation, absolute difference between ground truth and anchor have been taken instead of normalized difference between ground truth and anchor. Why is it so? Am I missing something? Because paper have suggested this in equation (2) in page num. 5.

ssd_tensorflow/model.py

Line 82 in f381dba

loc_loss = tf.reduce_sum(smooth_l1(pred_locs - true_locs), reduction_indices=2) * positives