Miguel Riem de Oliveira
Miguel Riem de Oliveira
> > However, I suggest you add a flag or some other way to use dataset_corrected.json optionally, while using dataset.json as a default. > > Now it searches the dataset...
I mean in zoom at 9h, or if you prefer, we can meet at 11h in LAR.
> However here the dataset is already defined in the `data.yml` Ah, ok. So if its defined there, why can't we say in the yaml if its the dataset.json or...
Just to say. I accept some of the way stuff is done in ATOM may not be the best. But we should always consider consistency as an important thing. Like...
Hi @manuelgitgomes, my opinion is that for now if you can skip this better. Even if the dataset for now does not contain the fisheye images.
Well the assumption is that we will have one row where the first columns says "Averages", right? I think we can go with that, mut it should perhaps be written...
I like this idea from @Kazadhum .
> Post calibration : https://github.com/lardemua/atom/assets/31831043/5298d189-0289-4443-a367-a5221c0455a2 Looks good. Errors are most of the times under 1 pixel.
In the video there's a number always changing, I think that's the average reprojection error for each frame, which is almost always under 1.0