Closed cheneyuwu closed 1 year ago
The RTR paper assumes access to the ground-truth relative transformation of the first frame for initialization. For a fair comparison, consider adding a note to the localization benchmark saying using the first frame's ground-truth is allowed.
The RTR paper assumes access to the ground-truth relative transformation of the first frame for initialization. For a fair comparison, consider adding a note to the localization benchmark saying using the first frame's ground-truth is allowed.