face.evoLVe
face.evoLVe copied to clipboard
batchsize for detect and align
I'd like to know whether the detection and location of human face can only be processed with a single image. How can we improve efficiency?