Adapt to Mast3r

[wrld]

Thanks for your nice work!

Since the method is built on the dust3r model, I wonder if the authors have tried to adapt their method to Mast3r which is an improved version of dust3r. If you have tried this, would you like to share the pre-trained model? Thanks!

[HengyiWang]

Hi @wrld, it would be quite interesting to see if spann3r can also achieve metric predictions as in mast3r. However, as I currently do not have access to training resources (the cluster I've been using is closed), it's unlikely we’ll be able to provide additional pre-trained weights in the near future:(