Tim Ohliger

This PR should be resolved by #5450 since the example was completely changed.

@jaggerwang Do you have any update on the migration of the semantic segmentation tool for point clouds to open source? Yesterday I got an email from basic.ai marketing stating: >...

First of, I don't think you have to call `vis.poll_events()`, `vis.update_renderer()` and ` vis.update_geometry(obj_mesh)` in the callback function. Calling `vis.run()` already does this for you: https://github.com/isl-org/Open3D/blob/f02e7d24ea115e716445a7fae5093bce60a37d20/cpp/open3d/visualization/visualizer/Visualizer.cpp#L274-L288 But you should return...

If you create a PR I think it would be faster to get implemented/reviewed/merged. Seems like a small but useful change.

I think this got fixed in #6874 and will be released with 0.19.0.

Probably the same. You could try using numpy 1.26.4, if your code and other dependencies allow it, to see if the segmentation fault still occures.

I split this from #6869 because it contains changes that actually modify the bindings a bit: Some C++ enums are exposed and the output of some `__repr__` methods changes. Also,...

Rebased since #6869 was merged. I moved the integration of stubs into the python package to #6917, since it is kind of separate and might require more investigation to have...

ToGoogleDocString strips any type annotations from function signatures, so every `docstring::ClassMethodDocInject(...)` call strips typing info. 703beadf9 fixes this, resulting in more types displayed in documentation, as well as, more types...

I do not know how to fix the last two errors right now, so for now I will have to leave them and maybe fix them later. Now I will...