Skip to content

[IROS2022] Project page for Scene-level tracking and reconstruction without object-prior.

Notifications You must be signed in to change notification settings

changhaonan/STAR-no-prior

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 

Repository files navigation

Scene Level Tracking And Reconstruction without object prior

Generate the geometry model, tracking the position and deformation for each visible individual objects in the scene without object prior.

More results on challenging scenes (RGB, Model, Segmentation)

Comparison with SurfelWarp (RGB, Ours, SurfelWarp)

Watch the differences between two methods in handling topological change between the mario and the hand.

Interactive demo (Still ongoing)

An interacitve is also provided for better evaluation of our results. Interactive demo

Source Code

For source code, please go to https://github.com/changhaonan/StarHub. Code is only tested on Ubuntu 20.04.

Cite this work

@inproceedings{chang2022scene,
  title={Scene-level Tracking and Reconstruction without Object Priors},
  author={Chang, Haonan and Boularias, Abdeslam},
  booktitle={2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages={3785--3792},
  year={2022},
  organization={IEEE}
}

About

[IROS2022] Project page for Scene-level tracking and reconstruction without object-prior.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published