I am facing a very odd problem that has to do with the reconstruction of the mesh and the rgb data itself. I am using the mp3d dataset in habtitat and want to build a semantic map with the semantic annotations of the navmesh. In the navmesh I found this room with a table but in the rgb image it does not exist? I would assume that mesh is constructed from the rgb and depth data. Did the navmesh computation possibly get confused by the pattern of the carpet?
The scene ID is 17DRP5sb8fy.
This is from the navmesh and contains a table object with position and dimension.
Hello,
I am facing a very odd problem that has to do with the reconstruction of the mesh and the rgb data itself. I am using the mp3d dataset in habtitat and want to build a semantic map with the semantic annotations of the navmesh. In the navmesh I found this room with a table but in the rgb image it does not exist? I would assume that mesh is constructed from the rgb and depth data. Did the navmesh computation possibly get confused by the pattern of the carpet?
The scene ID is 17DRP5sb8fy.
This is from the navmesh and contains a table object with position and dimension.
This is the rgb data.