app.py

import datetime
import gradio as gr
import numpy as np
import os

from paddlespeech.cli.tts.infer import TTSExecutor

#该函数有3个输入参数和2个输出参数
def greet(name,  file):

    print(file)

    #语音合成
    wavname = f"{datetime.datetime.now().strftime('%Y%m%d%H%M%S')}.wav"
    tts = TTSExecutor()
    tts(text=name, output=wavname)
    output_file = f"results/{datetime.datetime.now().strftime('%Y%m%d%H%M%S')}.mp4"

    #处理视频
    ckpt = os.path.join(os.path.dirname(__file__), "checkpoints/wav2lip.pth")
    audio = os.path.join(os.path.dirname(__file__), wavname)
    out =  os.path.join(os.path.dirname(__file__),  output_file)
    print(ckpt)
    print(audio)
    print(out)
    os.system(f"python3.10 ./inference.py --checkpoint_path {ckpt} --face_det_batch_size 1 --face {file} --audio {audio} --outfile={out}")
    # video_file = os.path.join(os.path.dirname(__file__), "result_voice5.mp4")
    # video_bytes = open(video_file, "rb").read()

    return wavname, out




demo = gr.Interface(
    fn=greet,
    #按照处理程序设置输入组件
    inputs=[gr.Text(placeholder="输入要转的文本"), gr.Image(type="filepath")],
    #按照处理程序设置输出组件
    outputs=[gr.Audio(), gr.Video(label="Processed Video")],
)

if __name__ == "__main__":
     demo.launch()