First, thanks for your sharing. But these days, I had been confused by the usage of result_pose that described in the paper with 27 parameters. Would you please to help me figure out the order of result_pose? I'm really want to verify the effect of this algorithm in the 3D space.