A question about quantitative surface normal estimation

[nickle-fang]

I appreciate the great work your team has done which gave us much inspiration! I have a small question about the evaluation of the normal estimation result in Table 2 in your paper: How did you select the data list of OmniObject3D when you compared the quantitative results with other methods? I wonder because I am comparing the different methods of their performance in different datasets. Could you please offer the list you used? Thanks a lot!

[fuxiao0719]

hi, @JUGGHM, can you kindly provide a response to this question?

[MiaApr]

I’m also curious about how the evaluation set of the OmniObject3D dataset was selected. Looking forward to your reply.