A question about the trajectory shown in Fig. 4

[zky72394-coder]

Dear Authors,

I hope this email finds you well.

I have been reading your paper with great interest, and I learned a lot from your work. I especially found the drift correction experiment in Fig. 4 very impressive.

I am currently trying to better understand and reproduce this part of the experiment. While reading the paper and going through the released code, I became a little confused about how the trajectories in Fig. 4 were obtained. In particular, I could not clearly find the implementation or script that generates the compared trajectories shown in the figure.

I was therefore wondering whether you might kindly clarify this point for me. If possible, I would be very grateful if you could let me know:
1. how the trajectories in Fig. 4 were computed,
2. which trajectories correspond to the poses before and after pose-graph optimization,
3. whether there were any additional post-processing or plotting steps that are not included in the public code.

I apologize if I have overlooked something in the repository. It is very possible that I missed the relevant part. I would sincerely appreciate any guidance you could offer.

Thank you very much for your time and for sharing your excellent work.

Best regards,
Ethan Carter