fungtion
fungtion
@apchenstu what's the parameter `img_fp.npy` like in sof/scripts/DataPreprocess.ipynb? And it seems we need this to generate cam2world.npy and intrinsics.npy
@dmalyavin you can try to replace tf stylegan2 with [pytorch stylegan2](https://github.com/NVlabs/stylegan2-ada-pytorch) .
it is about the source data training after the line 198, and the target data training above.
@xdnjust https://pan.baidu.com/s/1owUIJeHV2TRemGW4EVw-VQ 2f1m
@seuyou https://drive.google.com/file/d/1ylT-NufHgSu33_Se4QbEZzu2vU1qKkmC/view?usp=sharing
@JasonGUTU using stylegan_w+ for face is still far from ideal, can you show that how to split stylegan? I did that following PGGAN-Multi-Z, but it reported error that in x...
@ReverseSystem001 Can you show me where the hyperparameters m and s?
It's decieded not only by the number, but also the similarity between the source and target images, the bigger differences, the more data needed.
@HAOCHENYE 可以参考https://github.com/JDAI-CV/FaceX-Zoo/issues/18
change `Adam(lr=self.lrG*loss_config['lr_factor'], beta_1=0.5).get_updates(weightsGA,[], loss_GA)` to `Adam(lr=self.lrG*loss_config['lr_factor'], beta_1=0.5).get_updates(loss_GA, weightsGA)` may help to solve it.