Hi @yhyang-myron.
I found your work very interesting, and I believe the data curation pipeline described in Section 3.3 of your arXiv paper is a valuable contribution on its own.
Would it be possible to share more details about this part of the work, such as the code, data splits, or any implementation specifics? I'm particularly interested in the Objaverse subset of 180k high-quality 3D shapes and the steps involved in generating the final meshes (e.g., component merging, removal of non-visible faces, and UDF computation).
Best,
Marios.
Hi @yhyang-myron.
I found your work very interesting, and I believe the data curation pipeline described in Section 3.3 of your arXiv paper is a valuable contribution on its own.
Would it be possible to share more details about this part of the work, such as the code, data splits, or any implementation specifics? I'm particularly interested in the Objaverse subset of 180k high-quality 3D shapes and the steps involved in generating the final meshes (e.g., component merging, removal of non-visible faces, and UDF computation).
Best,
Marios.