question about 3dgs train

[twblaotang]

Thank you for your excellent work. I have a question about reconstruction.
I noticed that in your paper, you analyzed that there would be slight changes in the pose after using Viewcrafter compared to the input trajectory. When subsequently trainning 3DGS with the results of Viewcrafter, is it necessary to re-optimize the poses using Colmap or Dust3r, or is it okay to just use the original poses? Is there any difference ?