Different results on paper and youtube presentation

[maulanaazhari]

Hi, thanks for open sourcing the code.

I am curious about the sequence by sequence results on KITTI dataset and found the results on the youtube presentation you provided

however, the results looks different with the ones in the paper

I also have tried to run the code on kitti, but I got quite different results compared to the paper.

In any case you updated the experiments in the paper (compared with the youtube video), would you care to share the detailed results on the KITTI dataset?