app.py

import os
from PIL import Image
import torch
import gradio as gr
import torch
torch.backends.cudnn.benchmark = True
from torchvision import transforms, utils
from util import *
from PIL import Image
import math
import random
import numpy as np
from torch import nn, autograd, optim
from torch.nn import functional as F
from tqdm import tqdm
import lpips
from model import *
import urllib.request

#from e4e_projection import projection as e4e_projection

from copy import deepcopy
import imageio

import os
import sys
import numpy as np
from PIL import Image
import torch
import torchvision.transforms as transforms
from argparse import Namespace
from e4e.models.psp import pSp
from util import *
from huggingface_hub import hf_hub_download

device= 'cpu'
model_path_e = hf_hub_download(repo_id="aijack/e4e", filename="e4e.pt")
ckpt = torch.load(model_path_e, map_location='cpu')
opts = ckpt['opts']
opts['checkpoint_path'] = model_path_e
opts= Namespace(**opts)
net = pSp(opts, device).eval().to(device)
# Fetch image for analysis
img_url = "http://claireye.com.tw/img/230212a.jpg"
urllib.request.urlretrieve(img_url, "pose.jpg")
@ torch.no_grad()
def projection(img, name, device='cuda'):
 
    
    transform = transforms.Compose(
        [
            transforms.Resize(256),
            transforms.CenterCrop(256),
            transforms.ToTensor(),
            transforms.Normalize([0.5, 0.5, 0.5], [0.5, 0.5, 0.5]),
        ]
    )
    img = transform(img).unsqueeze(0).to(device)
    images, w_plus = net(img, randomize_noise=False, return_latents=True)
    result_file = {}
    result_file['latent'] = w_plus[0]
    torch.save(result_file, name)
    return w_plus[0]




device = 'cpu' 


latent_dim = 512

model_path_s = hf_hub_download(repo_id="aijack/stylegan2", filename="stylegan2.pt")
original_generator = Generator(1024, latent_dim, 8, 2).to(device)
ckpt = torch.load(model_path_s, map_location=lambda storage, loc: storage)
original_generator.load_state_dict(ckpt["g_ema"], strict=False)
mean_latent = original_generator.mean_latent(10000)

generatorjojo = deepcopy(original_generator)

generatorcaitlyn = deepcopy(original_generator)

generatorart = deepcopy(original_generator)

generatorsketch = deepcopy(original_generator)


transform = transforms.Compose(
    [
        transforms.Resize((1024, 1024)),
        transforms.ToTensor(),
        transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
    ]
)




modeljojo = hf_hub_download(repo_id="aijack/jojo", filename="jojo.pt")


ckptjojo = torch.load(modeljojo, map_location=lambda storage, loc: storage)
generatorjojo.load_state_dict(ckptjojo["g"], strict=False)

modelcaitlyn = hf_hub_download(repo_id="aijack/arcane", filename="arcane.pt")

ckptcaitlyn = torch.load(modelcaitlyn, map_location=lambda storage, loc: storage)
generatorcaitlyn.load_state_dict(ckptcaitlyn["g"], strict=False)

modelart = hf_hub_download(repo_id="aijack/art", filename="art.pt")

ckptart = torch.load(modelart, map_location=lambda storage, loc: storage)
generatorart.load_state_dict(ckptart["g"], strict=False)


modelSketch = hf_hub_download(repo_id="aijack/sketch", filename="sketch.pt")

ckptsketch = torch.load(modelSketch, map_location=lambda storage, loc: storage)
generatorsketch.load_state_dict(ckptsketch["g"], strict=False)

def inference(img, model):  
    img.save('out.jpg')  
    aligned_face = align_face('out.jpg')
        
    my_w = projection(aligned_face, "test.pt", device).unsqueeze(0)
    if model == 'JoJo':
        with torch.no_grad():
            my_sample = generatorjojo(my_w, input_is_latent=True)  
    elif model == 'Caitlyn':
        with torch.no_grad():
            my_sample = generatorcaitlyn(my_w, input_is_latent=True)
    elif model == 'Art':
        with torch.no_grad():
            my_sample = generatorart(my_w, input_is_latent=True)
    else:
        with torch.no_grad():
            my_sample = generatorsketch(my_w, input_is_latent=True)
            
    
    npimage = my_sample[0].permute(1, 2, 0).detach().numpy()
    imageio.imwrite('filename.jpeg', npimage)
    return 'filename.jpeg'
  
title = "JoJoGAN"
description = "Gradio Demo for JoJoGAN: One Shot Face Stylization. To use it, simply upload your image, or click one of the examples to load them. Read more at the links below."

article = "<p style='text-align: center'><a href='http://claireye.com.tw'>Claireye</a> | 2023</p>"

examples=[['pose.jpg','Art']]
gr.Interface(inference, [gr.inputs.Image(type="pil"),gr.inputs.Dropdown(choices=['JoJo', 'Caitlyn','Art','Sketch'], type="value", default='JoJo', label="Model")], gr.outputs.Image(type="numpy"),title=title,description=description,article=article,allow_flagging=False,examples=examples,allow_screenshot=False).launch()