app.py

# -* coding:UTF-8 -*
# !/usr/bin/env python
import numpy as np
import gradio as gr
import roop.globals
from roop.core import (
    start,
    decode_execution_providers,
)
from roop.processors.frame.core import get_frame_processors_modules
from roop.utilities import normalize_output_path
import os
from PIL import Image
import spaces
import uuid
import onnxruntime as ort

@spaces.GPU
def swap_face(source_file, target_file, doFaceEnhancer):
    session_id = str(uuid.uuid4())  # Tạo một UUID duy nhất cho mỗi phiên làm việc
    session_dir = f"temp/{session_id}"
    os.makedirs(session_dir, exist_ok=True)

    source_path = os.path.join(session_dir, "input.jpg")
    target_path = os.path.join(session_dir, "target.jpg")

    source_image = Image.fromarray(source_file)
    source_image.save(source_path)
    target_image = Image.fromarray(target_file)
    target_image.save(target_path)

    print("source_path: ", source_path)
    print("target_path: ", target_path)

    output_path = os.path.join(session_dir, "output.jpg")
    normalized_output_path = normalize_output_path(source_path, target_path, output_path)

    frame_processors = ["face_swapper", "face_enhancer"] if doFaceEnhancer else ["face_swapper"]
    headless = True
    keep_fps = True
    keep_audio = True
    keep_frames = False
    many_faces = False
    video_encoder = "libx264"
    video_quality = 18
    max_memory = "4G"
    execution_providers = ["CUDAExecutionProvider", "CPUExecutionProvider"]  # Ưu tiên GPU, sau đó là CPU
    execution_threads = 4
    reference_face_position = 0
    similar_face_distance = 0.6

    print("Available providers:", ort.get_available_providers())  # Kiểm tra các provider hiện có
    print("Configured execution providers:", execution_providers)

    for frame_processor in get_frame_processors_modules(frame_processors):
        if not frame_processor.pre_check():
            print(f"Pre-check failed for {frame_processor}")
            return

    roop.globals.source_path = source_path
    roop.globals.target_path = target_path
    roop.globals.output_path = normalized_output_path
    roop.globals.frame_processors = frame_processors
    roop.globals.headless = headless
    roop.globals.keep_fps = keep_fps
    roop.globals.keep_audio = keep_audio
    roop.globals.keep_frames = keep_frames
    roop.globals.many_faces = many_faces
    roop.globals.video_encoder = video_encoder
    roop.globals.video_quality = video_quality
    roop.globals.max_memory = max_memory
    roop.globals.execution_providers = execution_providers
    roop.globals.execution_threads = execution_threads
    roop.globals.reference_face_position = reference_face_position
    roop.globals.similar_face_distance = similar_face_distance

    start()
    return output_path

app = gr.Interface(
    fn=swap_face, 
    inputs=[
        gr.Image(), 
        gr.Image(), 
        gr.Checkbox(label="Face Enhancer?", info="Do face enhancement?")
    ], 
    outputs="image"
)
app.launch()