Hello, im trying to put my girlfriend's face into a tv series scene. I want to ask what is the fastest configuration, or maybe you can put some option/preset for regular(and maybe low-end pc) users who just wants to do 'not so important' stuff.
For example; both my models and video scene, they are all high quality, i think training will be so accurate i have no doubt about it but i dont want to make something so professional you know, its ok just to change faces and catch the right colour tone. when i do train with original trainer(i know about lightweight still i choose original), i have like 80k iterations still the model is blurry when i convert it. well i can do 150k.. or 300k or more its alright i get it the more i get the better it will be but i think there has to be some faster way than that, i mean it feels like it shouldnt be so hard to swap faces when you got your alignments then just replace faces and fix the borders or am i wrong?
I mean it could be more useful if people can do low quality swaps to videos. By low quality, not blurry but only just replace faces instead of blurring it maybe..? And since english isnt my primary language, sorry if i say something rude or like that.