import torch
from torchvision.transforms.functional import normalize

from models import CodeFormer
from utils import imwrite, img2tensor, tensor2img
from facelib.utils.face_restoration_helper import FaceRestoreHelper
from huggingface_hub import hf_hub_download

REPO_ID = "leonelhs/gfpgan"

pretrain_model_path = hf_hub_download(repo_id=REPO_ID, filename="CodeFormer.pth")

if __name__ == '__main__':

    device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

    net = CodeFormer(dim_embd=512, codebook_size=1024, n_head=8, n_layers=9,
                                          connect_list=['32', '64', '128', '256']).to(device)

    checkpoint = torch.load(pretrain_model_path)['params_ema']
    net.load_state_dict(checkpoint)
    net.eval()


    face_helper = FaceRestoreHelper(
        upscale_factor=2,
        face_size=512,
        crop_ratio=(1, 1),
        det_model='retinaface_resnet50',
        save_ext='png',
        use_parse=True,
        device=device)

    input_img_list = ["/home/leonel/Pictures/lowres13.jpg"]

    # -------------------- start to processing ---------------------
    for i, img_path in enumerate(input_img_list):
        # clean all the intermediate results to process the next image
        face_helper.clean_all()
        img = img_path

        face_helper.read_image(img)
        # get face landmarks for each face
        num_det_faces = face_helper.get_face_landmarks_5(
            only_center_face=False, resize=640, eye_dist_threshold=5)
        print(f'\tdetect {num_det_faces} faces')
        # align and warp each face
        face_helper.align_warp_face()

        # face restoration for each cropped face
        for idx, cropped_face in enumerate(face_helper.cropped_faces):
            # prepare data
            cropped_face_t = img2tensor(cropped_face / 255., bgr2rgb=True, float32=True)
            normalize(cropped_face_t, (0.5, 0.5, 0.5), (0.5, 0.5, 0.5), inplace=True)
            cropped_face_t = cropped_face_t.unsqueeze(0).to(device)

            try:
                with torch.no_grad():
                    output = net(cropped_face_t, w=0.5, adain=True)[0]
                    restored_face = tensor2img(output, rgb2bgr=True, min_max=(-1, 1))
                del output
                torch.cuda.empty_cache()
            except Exception as error:
                print(f'\tFailed inference for CodeFormer: {error}')
                restored_face = tensor2img(cropped_face_t, rgb2bgr=True, min_max=(-1, 1))

            restored_face = restored_face.astype('uint8')
            face_helper.add_restored_face(restored_face, cropped_face)

        # paste_back
        has_aligned = False
        suffix = None
        if not has_aligned:
            bg_img = None
            face_helper.get_inverse_affine(None)
            restored_img = face_helper.paste_faces_to_input_image(upsample_img=bg_img, draw_box=False)
            imwrite(restored_img, "pretty.png")