6-DOF pose estimation (SnapML)
Dear Support,
First of all, congrats on the new 3.0 update! What an awesome step forward, just amazing. The possibilities of SnapML are something else.
Say i would like to train my own model that lets me detect the position and orientation of a physical object (6-DOF tracking). Sort of like a 3D model tracker, be it a piece of packaging, a book or a toy car.
I was wondering if you had some tips/links to tensorflow/PyTorch/other approaches to get me started in the right direction. Are there models available that let me do this myself?
Hi Tom Geurts,
Training a 6-DOF tracker involves many complications and we believe it's beyond the scope of our current ML tutorials. However, we will continue to provide support and may release more tutorials to cover more advanced topics including 3D tracking in the future.
Best,
Eric