Quantitative Results of Single-view Reconstruction

yccyenchicheng / AutoSDF

233 stars 29 forks source link

Quantitative Results of Single-view Reconstruction #9

Closed Silverster98 closed 2 years ago

Silverster98 commented 2 years ago

I have noticed that the IoU of Pix2Vox on ShapeNet is reported as 0.467 in your paper, but the value reported in the original paper is over 0.6. I don't find too much difference in the setting between AutoSDF and Pix2Vox. Can you explain more?

yccyenchicheng commented 2 years ago

Hi,

We use the split provided in DISN so the number is different.

Thanks!