Weird results using pretrained model

[zhuhaozh]

Using the inference.py, got some weird results on my images.

[eshuka]

I have a same problem. did you fix it?

[Torreto666]

I think the loss function has some problems

[changfali]

Got that problem,too. Do you have some ideas? @Torreto666 @eshuka @tomguluson92

[punitha-valli]

@zhuhaozh @Torreto666 @changfali @eshuka @tomguluson92

How do you get the depth map, Can you please clear it Hope for a reply

Thanks in advance

[Jason0917]

Same problem here. Any thoughts on this issue? XD

[tomguluson92]

Hi, everyone, the current model here is only trained on a few faces with extreme poses, so the model is not generalizing well to the common situations.

Train this PRNET with full dataset is all you need :)

[xuncheng1997]

Hi, everyone, the current model here is only trained on a few faces with extreme poses, so the model is not generalizing well to the common situations. Train this PRNET with full dataset is all you need :)

When I train on full datasets, the results are still not good. How can i solve it.

[shiwen-cheng]

I have the same problem, has there anyone solved this? please! i really need help :)