Hi, this is one issue I mentioned in the previous block. I bring it up again just in case.
My team could not reproduce the result of this algorithm. The trajectory plot just froze and stopped working halfway. I think the reason is that there's tracking loss caused by lack of features, and the algorithm doesn't have a re-initialization process (didn't see that in the paper). How could I solve that? If I need to adjust some parameters to reproduce the result, could you provide me the parameters used?
Hi, this is one issue I mentioned in the previous block. I bring it up again just in case. My team could not reproduce the result of this algorithm. The trajectory plot just froze and stopped working halfway. I think the reason is that there's tracking loss caused by lack of features, and the algorithm doesn't have a re-initialization process (didn't see that in the paper). How could I solve that? If I need to adjust some parameters to reproduce the result, could you provide me the parameters used?