Congratulations on your work, and thanks for sharing, I am a big fan.
I checked that you cannot currently publish any fitting models, however I am using your model to fit it to images via inverse rendering (using redner, "Differentiable Monte Carlo Ray Tracing through Edge Sampling").
I am using pytorch and noticed that sometimes (not always which is weird) if you run torch.isfinite().all() on the loaded model uvs returns False, resulting to an exception in redner.
Have you ever experienced something similar, or can think of why this can happen?
Hi,
Congratulations on your work, and thanks for sharing, I am a big fan. I checked that you cannot currently publish any fitting models, however I am using your model to fit it to images via inverse rendering (using redner, "Differentiable Monte Carlo Ray Tracing through Edge Sampling").
I am using pytorch and noticed that sometimes (not always which is weird) if you run
torch.isfinite().all()
on the loaded model uvs returnsFalse
, resulting to an exception in redner.Have you ever experienced something similar, or can think of why this can happen?
Thank you again, Antonis