search

xiesc / slamKitti_noros

Apply LOAM on KITTI
10 stars 5 forks source link

kitti experiment #2

Open ghost opened 5 years ago

ghost commented 5 years ago

Thank you for providing your source code. Do you get the same result on kitti dataset as zhang ji's paper using your code?

xiesc commented 5 years ago

This commit is a little out-date than my current version. I have tested it on the test sequence with some tricks like obstacle removing e.g.. Finally the loss of it can achieve about 1.00%. There are some works still need to do to achieve Zhang's results, since there are a lot of params can be tuned.