You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
GPU is only for the 2D line detection from the video stream. Currently, the state-of-the-art 2D line detection methods are based on deep learning, which can guarantee the best performance in both accuracy and speed. For the 2D-3D pose tracking, GPU is not necessary. You can try other non-learning 2D line detection method to replace the afm package (such as LSD, TPAMI 2008), it can run in realtime but add outlier 2D-3D correspondences.
@levenberg Thanks for your quick answering. I'll try to use other 2D line detection method to run the project on CPU. I'll get back for your reference when meeting some issue.
As I see afm using cuda, so is GPU mandatory to run this project. Is there a way to run it on CPU?
Thanks
The text was updated successfully, but these errors were encountered: