Thanks for sharing the paper and code. It's a great work!
Since adding corruptions to the original datasets will variant by different persons, for fair comparison, could you also share the corrupted versions of ScanObjectNN-C and ShapeNet-C? I think it will also greatly facilitate much more followings of your work.
Thanks for sharing the paper and code. It's a great work!
Since adding corruptions to the original datasets will variant by different persons, for fair comparison, could you also share the corrupted versions of ScanObjectNN-C and ShapeNet-C? I think it will also greatly facilitate much more followings of your work.
Thanks. Looking forward to you reply.