Where can I replace the VGG model with ResNet model ?

[t-morik]

Hello, I always enjoy reading your papers.

I am currently using your Pix2Vox as a trial study, and I was wondering where I can change the "we replace VGG with ResNet as the new backbone network" mentioned in the follow-up work Pix2Vox++ paper.

I guess the relevant part is line 18 in encoder.py, Is it correct to assume that the code here is still in the VGG network?

vgg16_bn = torchvision.models.vgg16_bn(pretrained=True)

I've just recently started to challenge machine learning from the bioengineering field, so I'm sorry if this is a misguided question

Thank you very much.

[hzxie]

Yes, you are right.
BTW, the code for Pix2Vox++ is open-sourced at https://gitlab.com/hzxie/Pix2Vox.

[t-morik]

Thank you !