sair-lab / iSLAM

iSLAM: Imperative SLAM (RA-L 2024) is a novel Visual-Inertial SLAM using Self-supervised Learning
https://sairlab.org/iSLAM/
Other
109 stars 9 forks source link

How to visualize the dense reconstruction presented in your paper? #3

Open qian5683 opened 3 months ago

qian5683 commented 3 months ago

Thanks for your good work! I want to know how to visualize the reconstructed map

image
FuTaimeng commented 1 month ago

Hi, thank you for the question.

We used a straightforward mapping method, as the localization process doesn't rely on this map: we project each left image into 3D space using stereo depth, transform the point clouds with the predicted poses, and then aggregate all frames together.

I've uploaded the code as mapper.py in the dev branch, hope you find it helpful.