Thanks for the interesting work. I am curious how well does the model scale with the number of points. From the paper I see, that the model can predict ~16k points for shape completion. Will the model be able to scale if one wants to predict 100K points or 1M points? Are there any particular issues which it can run into, like memory issues?
Thanks for the interesting work. I am curious how well does the model scale with the number of points. From the paper I see, that the model can predict ~16k points for shape completion. Will the model be able to scale if one wants to predict 100K points or 1M points? Are there any particular issues which it can run into, like memory issues?