yuxng / PoseCNN

A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes
https://rse-lab.cs.washington.edu/projects/posecnn/
MIT License
761 stars 245 forks source link

What global optimization step was taken? #56

Open Kaju-Bubanja opened 6 years ago

Kaju-Bubanja commented 6 years ago

I'm trying to record my own dataset and I'm using a motion capture system. Now I was wondering how you did the camera tracking? Did you also use a motion capture system or some form of visual slam? Also I'm wondering what optimization you performed at the end, because I heard the camera trajectory might be quite unreliable without optimization. Is there maybe an appendix to your impressive paper, where some of the details are explained in more depth?