Dear,
I ran the eval.py using the pre-trained weights of the loop dataset (named/RobotCar_loop_mapnet_mapnet_learn_beta_learn_gamma/epoch_300_.pth.tar').
However, the result didn't match that from the paper (see figure below).

May I have some suggestions for it?
Best!
Jialu
Dear,

I ran the eval.py using the pre-trained weights of the loop dataset (named/RobotCar_loop_mapnet_mapnet_learn_beta_learn_gamma/epoch_300_.pth.tar').
However, the result didn't match that from the paper (see figure below).
May I have some suggestions for it?
Best!
Jialu