app.py

import gradio as gr

import argparse
import functools

import numpy as np
import torch
from infer_contrast import run

from utils.reader import load_audio
from utils.utility import add_arguments, print_arguments

def voiceRecognition(audio1,audio2):
    text = run(audio1,audio2)
    return text


title = "Voice Recognition"
description = "This voice recognition demo(Chinese Format) is a simple implementation based on ResNet. It used ArcFace Loss and an open source Chinese voice corpus - zhvoice."

inputs = [gr.inputs.Audio(source="upload",type="filepath", label="Speaker1"),
          gr.inputs.Audio(source="upload",type="filepath", label="Speaker2")]
          
article = (
    "<p style='text-align: center'>"
    "<a href='https://github.com/yeyupiaoling/VoiceprintRecognition-Pytorch' target='_blank'>💻 Original Dataset</a> | "
    "<a href='https://github.com/fighting41love/zhvoice' target='_blank'>🎙️ zhvoice dataset</a> | "
    "</p>"
)

examples = [
    ["samples/李云龙1.wav", "samples/李云龙2.wav"],
    ["samples/马保国1.wav", "samples/马保国2.wav"],
    ["samples/周杰伦1.wav", "samples/周杰伦2.wav"]]

interface = gr.Interface(
    fn=voiceRecognition,
    inputs=inputs,
    outputs="text",
    title=title,
    description=description,
    examples=examples,
    enable_queue=True)
interface.launch(debug=True,share=True)