some problem with KITTI's mean RTE

[callmewanzi]

Thanks for your job! when i did some work based on your models, I found that the total RTE after tested all kitti_dataset was not equal to the average RTE number i counted(use each scan pair to count a mean number). and i checked the coding ,found the problem was that :kitti experiments didn't clean the SummaryBoard: def reset_meter(self, name): self.meter_dict[name].reset() That was i thought, If it was correct,please let me know. thans for your work again