third_party/GPEN/segmentation2face.py

'''
@paper: GAN Prior Embedded Network for Blind Face Restoration in the Wild (CVPR2021)
@author: yangxy ([email protected])
'''
import os
import cv2
import glob
import time
import numpy as np
from PIL import Image
import __init_paths
from face_model.face_gan import FaceGAN

class Segmentation2Face(object):
    def __init__(self, base_dir='./', size=1024, model=None, channel_multiplier=2, narrow=1, is_norm=True):
        self.facegan = FaceGAN(base_dir, size, model, channel_multiplier, narrow, is_norm)

    # make sure the face image is well aligned. Please refer to face_enhancement.py
    def process(self, segf):
        # from segmentations to faces
        out = self.facegan.process(segf)

        return out
        

if __name__=='__main__':
    model = {'name':'GPEN-Seg2face-512', 'size':512}
    
    indir = 'examples/segs'
    outdir = 'examples/outs-seg2face'
    os.makedirs(outdir, exist_ok=True)

    seg2face = Segmentation2Face(size=model['size'], model=model['name'], channel_multiplier=2, is_norm=False)

    files = sorted(glob.glob(os.path.join(indir, '*.*g')))
    for n, file in enumerate(files[:]):
        filename = os.path.basename(file)
        
        segf = cv2.imread(file, cv2.IMREAD_COLOR)

        realf = seg2face.process(segf)
        
        segf = cv2.resize(segf, realf.shape[:2])
        cv2.imwrite(os.path.join(outdir, '.'.join(filename.split('.')[:-1])+'.jpg'), np.hstack((segf, realf)))
        
        if n%10==0: print(n, file)