abdallahdib

thanks Raeen for your quick prompt. after following ur suggestion everything works just fine! Many thanks!!

at line 87 from the script [here](https://github.com/zhhoper/DPR/blob/master/testNetwork_demo_1024.py) u can see that network estimates in addition to the final image output, the light (sh) of the input image (outputSH) outputImg, _,...

i share here a dirty and messy code i did urgently for a deadline. these 3 scripts allow to prepare data for styleflow. azureRun.py : used to fetch images attributes...

Hey Rameen, thanks for ur response, i meant if we can get the original images and not the latent code . is it possible? are images from fig 10 not...

yes clear. thank you so much!

uvParametrization.pickle contains the barycentric coordinates used to generate the texture map from the per vertex color. Flame already have these barycentric coordinates which can be found here: https://github.com/TimoBolkart/BFM_to_FLAME/blob/main/col_to_tex.py these coordinates...

Hi, are u using only the landmark loss for training? what are ur learning rates? does the loss converges? can u show some intermediate output? please provide more details to...

Hey, u should take note that in the paper we were using basel 2009 where here u are using basel 2017. ur Regularizations terms need to be revisited and adjusted....

Ok joining back this discussion :) Here some suggestions that i hope will help: 1. Please use celeba cropped and aligned images (img_align_celeba_png.7z) 2. enfore landmarks to be of the...

just a simple initialization of the layer (set weights to 0 and bias to a constant is enough)