search

lmb-freiburg / deeptam

DeepTAM: Deep Tracking and Mapping https://lmb.informatik.uni-freiburg.de/people/zhouh/deeptam/
GNU General Public License v3.0
233 stars 41 forks source link

Tracking code with given weights give very high RMSE #4

Closed stillbreeze closed 5 years ago

stillbreeze commented 5 years ago

Hi,

I tried running the tracking code (deeptam/tracking/examples/example_advanced_sequence.py) but the translational RMSE printed at the end is 679702 m/s. The visualization shows no current image. The weights are loading fine. Any idea what might be going wrong? Here's the output for one of the intermediate frames. Also, every time I run it, there is a slight variation in the RMSE values. Shouldn't it remain the same since we are loading the pre trained weights. Thanks for the help in advance!

track_res_167

image

JingyuChan commented 5 years ago

@stillbreeze I have met the same problem and I just fixed it. A possible solution is to change the skimage version to 0.14.2 if your skimage version is higher than 0.15.0. Hope that can help you a little bit.

stillbreeze commented 5 years ago

Thanks @JingyuChan ! It works fine now.

zhz218 commented 5 years ago

I fixed the problems in the latest commit. Now it should also work with scikit-image 0.15.x. Thanks for pointing this out.