search

Harry-Zhi / semantic_nerf

The implementation of "In-Place Scene Labelling and Understanding with Implicit Scene Representation" [ICCV 2021].
Other
418 stars 54 forks source link

About camera poses #28

Closed donjiaking closed 1 year ago

donjiaking commented 1 year ago

Hi, I wonder what is the camera format in traj_w_c.txt ? Does the filename "*_w_c" suggests it is in world to camera format? Thank you very much.

Harry-Zhi commented 1 year ago

Hi @donjiaking ,

The camera pose are 'camera to world' transformations (T_ab means transformation for b to a). Sorry for the confusion.

Please note that the camera coordinates follow OpenCV convention where x,y,z axes are 'right', 'down', 'forward'.