Original code could not adopt TUM rgbd dataset conveniently.
Therefore, I add -tum parameter for loading first pose from TUM groundtruth.txt.
With first pose, the output *.poses file could fit with groundtruth.txt and compare ate with evaluate_ate.py provided by TUM.
Original code could not adopt TUM rgbd dataset conveniently. Therefore, I add -tum parameter for loading first pose from TUM groundtruth.txt. With first pose, the output *.poses file could fit with groundtruth.txt and compare ate with
evaluate_ate.pyprovided by TUM.