Got different result from the groundtruth

[lulu93]

Thanks for the code. I used the freiburg1_desk dataset to test your program, but got a very different result from the groudtruth. What should I do? Make a adjustment of the parameters of do something else? result: groudtruth:

[mp3guy]

When you run the evaluation tool from TUM it aligns the trajectories, they don't have to be in the same world frame of reference.

[lulu93]

why the numbers in fast.poses are so small? What is the unit of them?