Thank you for your excellent work! I tried applying the same approach to our pipeline, but I noticed that extracting SAM features takes a considerable amount of time, especially since we have six multi-view images. Is there a way to preprocess the SAM features before passing them to the 3D network?
Thanks
Thank you for your excellent work! I tried applying the same approach to our pipeline, but I noticed that extracting SAM features takes a considerable amount of time, especially since we have six multi-view images. Is there a way to preprocess the SAM features before passing them to the 3D network?
Thanks