Forward Direction in inverse kinematic

[huangsheng95]

hello! thanks for your great job! I have some question about the forward direction in inverse kinematics。

The forward direction is calculated by four joints (r_hip, l_hip, sdr_r, sdr_l) and their indices are defined in motion_representation.ipynb by

    # Face direction, r_hip, l_hip, sdr_r, sdr_l
    face_joint_indx = [2, 1, 17, 16]
    # l_hip, r_hip
    r_hip, l_hip = 2, 1

but in inverse kinematic，the index of r_hip and l_hip is swapped

    '''Get Forward Direction'''
    l_hip, r_hip, sdr_r, sdr_l = face_joint_idx

I‘m not sure how much it affects the motion representation process and results

[EricGuo5513]

Hi, thanks for letting me know. The order is not right. But fortunately, this should not affect the results, because the forward direction is obtained by the cross product of y-axis and any vector in upper body plane. Also, as long as you are using the forward kinematics function, the results should be always recovered.

huangsheng95 @.***> 于2023年4月25日周二 21:10写道：

hello! thanks for your great job! I have some question about the forward direction in inverse kinematics。

The forward direction is calculated by four joints (r_hip, l_hip, sdr_r, sdr_l) and their indices are defined in motion_representation.ipynb by

# Face direction, r_hip, l_hip, sdr_r, sdr_l
face_joint_indx = [2, 1, 17, 16]
# l_hip, r_hip
r_hip, l_hip = 2, 1

but in inverse kinematic，the index of r_hip and l_hip is swapped

'''Get Forward Direction'''
l_hip, r_hip, sdr_r, sdr_l = face_joint_idx

I‘m sure how much it affects the motion representation process and results

— Reply to this email directly, view it on GitHub https://github.com/EricGuo5513/HumanML3D/issues/55, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKRYNB4COXGRESU57SXA2V3XDCGZNANCNFSM6AAAAAAXL2GGTM . You are receiving this because you are subscribed to this thread.Message ID: @.***>