Dominik Winkelbauer

Results 157 comments of Dominik Winkelbauer

This is currently not implemented in BlenderProc, however it should be possible by slightly adjusting the Front3DLoader. The furniture objects are already loaded room-by-room: https://github.com/DLR-RM/BlenderProc/blob/main/blenderproc/python/loader/Front3DLoader.py#L407 So here you just need...

This might be a question to ask at https://github.com/thodan/bop_toolkit

Hey @Bainily028, in the coco annotations you find the `image_id` and `category_id` (=object id). In the scene_gt.json, the index should correspond the image id and you find the respective `obj_id`....

Hey @TheMr33, looks good, feel free to open a pull request. However I would not add sensor size to the `set_intrinsics_from_K_matrix` function, as it does not really make sense there.

This is probably an issue with blender on your system. Can you run the blender standalone gui?

Can you try whether you can run a more recent blender version when manually downloading it? https://download.blender.org/release/Blender4.2/blender-4.2.1-linux-x64.tar.xz

Hey @rb4005, I cannot really reproduce your problem. When running your code, the memory increased only for about 20MB per 1000 iterations. To be fair I added the following lines...

Does the addon work in blender 4.2.1 (without blenderproc)?

Hey @kamdkslansgka, yes that is possible using the bop writer. See for example here: https://github.com/DLR-RM/BlenderProc/blob/main/examples/basics/camera_object_pose/main.py#L61 The bop writer will also output masks for each object without occlusions.