search

TaatiTeam / MotionAGFormer

Official implementation of the paper "MotionAGFormer: Enhancing 3D Pose Estimation with a Transformer-GCNFormer Network" (WACV 2024).
Apache License 2.0
128 stars 15 forks source link

why rot = [0.1407056450843811, -0.1500701755285263, -0.755240797996521, 0.6223280429840088] ？ #13

Closed APeiZou closed 7 months ago

APeiZou commented 7 months ago

@SoroushMehraban @siarez hello, this is a great work，What I don't understand is that: why rot = [0.1407056450843811, -0.1500701755285263, -0.755240797996521, 0.6223280429840088] in demo.py and why are some values negative and where are their origins？

SoroushMehraban commented 7 months ago

Hi @APeiZou, I used MHFormer's demo and modified it to work with MotionAGFormer. It doesn't really matter I guess to follow whatever it's done in the for loop before plotting it. I don't think it's efficient either because it stores each frame as an image file and then tries to read and turn them into a video. Surely it can be done in a better way to take less time.