quat2mat

[ZhiyiHe1997]

sc_depth_pl/losses/inverse_warp.py

Line 63 in 2f76ce8

rotMat = torch.stack([w2 + x2 - y2 - z2, 2*xy - 2*wz, 2*wy + 2*xz,

if we use the quaternion coefficients as shown in image1

according to the Rodriguez formula，We can deduce the rotation matrix，as shown in image2

i find there are difference between formula and codes. Could you tell me the details of the changes you have made?

[JiawangBian]

Just copied from https://github.com/ClementPinard/SfmLearner-Pytorch/blob/master/inverse_warp.py. Didn't use it.

[ZhiyiHe1997]

yes, i saw it, may i ask it will influence the results?
have you think adout this trick?

[JiawangBian]

You can try this one "https://kornia.readthedocs.io/en/latest/_modules/kornia/geometry/conversions.html#quaternion_to_rotation_matrix" if you need it.

[JiawangBian]

yes, i saw it, may i ask it will influence the results? have you think adout this trick?

I tried it before and got very similar results compared to using euler angle on KITTI.

[ZhiyiHe1997]

ok , thanks for your reply. i will try it later.
Best regards.