search

PRBonn / SHINE_mapping

🌟 SHINE-Mapping: Large-Scale 3D Mapping Using Sparse Hierarchical Implicit Neural Representations (ICRA 2023)
MIT License
448 stars 32 forks source link

issue #37

Closed whwh747 closed 2 months ago

whwh747 commented 2 months ago

Thank you for your excellent work. I used kiss icp + python shine_batch.py config/maicity/maicity_batch.yaml and got some strange results. What could be the reason for this? 84781ce4bcf099b386576ace48fd340

YuePanEdward commented 2 months ago

Hi, I guess the reason is that the pose estimated by KISS-ICP on this sequence using the default parameter is not correct.

shine_issue37

As I also mentioned in this issue https://github.com/PRBonn/PIN_SLAM/issues/28, the simulation of the poses in MaiCity does not obey the common kinematics of robots. You may check here for more details: https://github.com/PRBonn/kiss-icp/issues/62. Mai-city was not designed to benchmark odometry systems.

whwh747 commented 2 months ago

thank you，I see, perhaps we can provide the poses obtained from another odometry to SHINE.