is the relative pose computation correct?

[ziming-liu]

    groundtruth_sequence[1:] = groundtruth_sequence[1:] - groundtruth_sequence[0] # get relative pose w.r.t. the first frame in the sequence 

I saw you got the relative pose by minus operation, but shouldn't we got the relative pose by times operation of two absolute pose?

such as 2T1 = inv(wT2) dot (wT1)

[ztysdu]

I am also confused about this problem before.But we can look at test.py
# use only last predicted pose in the following prediction last_pose = predict_pose_seq[-1]#[6] for i in range(len(last_pose)): last_pose[i] += answer[-1][i] # normalize angle to -Pi...Pi over y axis last_pose[0] = (last_pose[0] + np.pi) % (2 * np.pi) - np.pi answer.append(last_pose.tolist()) # print("answerlen",len(answer))
ground truth and the output of the model are consistent.So,it is right.