Possibly wrong calculation of ATE?

[maaft]

When testing the pose prediction accuracy (in pose_evaluation_utils.py), the ATE of a snippet is divided by the snippet length. Both predicted and GT snippets have the Identity transformation as their first entry. Therefore, the error on that first "pose" will be 0 which results in an overall snippet ATE that is considerably lower than it should be.

[warmkid]

I think the snippet way has some problem. the groudtruth pose divided into de snippet files has some problem caused the ate has some error,so the trajectory plot has some problem. But I think the network is good to learn the pose between the frames, the result has great room to improvement.