facenet-pytorch
facenet-pytorch copied to clipboard
Image normnalization or preprocessing
Hi first of all thanks for your work!
I am developing an app where I take faces and try to find the nearest face from a custom dataset. This custom dataset has a lot of variance in colors and I am getting poor results finding nearest faces (not so similar...). Maybe I need to preprocess the images from both domains first but I not find nothing in this repo. I am only using InceptionResnetV1(pretrained='vggface2').eval(), not the MTCNN. Thanks again!