I stumbled upon this github a while ago and this is truly a wonderful project. I am currently training my own stylegan on people. The problem I have when I want to use your google colab is that my photos often have backgrounds or don't portray the full person what results in a bad result. I was wondering if it was possible to train this on your own dataset of photos. This would help me so much and get my masters project for school in order.
I stumbled upon this github a while ago and this is truly a wonderful project. I am currently training my own stylegan on people. The problem I have when I want to use your google colab is that my photos often have backgrounds or don't portray the full person what results in a bad result. I was wondering if it was possible to train this on your own dataset of photos. This would help me so much and get my masters project for school in order.