app.py

import time
import streamlit as st
# from transformers import pipeline
import os
import torch
import datetime
import numpy as np
import soundfile
from wavmark.utils import file_reader
from audioseal import AudioSeal
import torchaudio
from pydub import AudioSegment
import io
import librosa
import ffmpeg

#from torchaudio.io import CodecConfig
# import numpy

# def my_read_file(audio_path, max_second):
#     signal, sr, audio_length_second = read_as_single_channel_16k(audio_path, default_sr)
#     if audio_length_second > max_second:
#         signal = signal[0:default_sr * max_second]
#         audio_length_second = max_second

#     return signal, sr, audio_length_second

def create_default_value():
    if "def_value" not in st.session_state:
        def_val_npy = np.random.choice([0, 1], size=32 - len_start_bit)
        def_val_str = "".join([str(i) for i in def_val_npy])
        st.session_state.def_value = def_val_str

def download_sample_audio():
    url = "https://keithito.com/LJ-Speech-Dataset/LJ037-0171.wav"
    with open("test.wav", "wb") as f:
        resp = urllib.request.urlopen(url)
        f.write(resp.read())
    
    wav, sample_rate = torchaudio.load("test.wav")
    return wav, sample_rate

# Main web app
def main():
    create_default_value()

    # st.title("MDS07 Demo Presentation")
    # st.write("https://github.com/wavmark/wavmark")
    markdown_text = """
    # MDS07 Demo Presentation
    [AudioSeal](https://github.com/jcha0155/audioseal) is the next-generation watermarking tool driven by AI. 
    You can upload an audio file and encode a custom 16-bit watermark or perform decoding from a watermarked audio.
    
    This page is for demonstration usage and only process **the first minute** of the audio. 
    If you have longer files for processing, we recommend using [our python toolkit](https://github.com/jcha0155/audioseal).
    """

    # 使用st.markdown渲染Markdown文本
    st.markdown(markdown_text)

    audio_file = st.file_uploader("Upload Audio", type=["wav", "mp3"], accept_multiple_files=False)

    if audio_file:
        # 保存文件到本地：
        # tmp_input_audio_file = os.path.join("/tmp/", audio_file.name)
        # st.markdown(tmp_input_audio_file)
        # with open(tmp_input_audio_file, "wb") as f:
        #     f.write(audio_file.getbuffer())
        # st.audio(tmp_input_audio_file, format="mp3/wav")


        #1st attempt
        #audio_path = " audio_file.name"

        # audio, sr = torchaudio.load(audio_file)
        # st.audio(audio_file, format="audio/mpeg")
        # audio= audio.unsqueeze(0)

        # st.markdown("SR")
        # st.markdown(sr)
        # st.markdown("after unsqueeze wav or mp3")
        # st.markdown(audio)
        
        #2nd attempt
        # Save file to local storage
        tmp_input_audio_file = os.path.join("/tmp/", audio_file.name)
        file_extension = os.path.splitext(tmp_input_audio_file)[1].lower()
        #st.markdown(file_extension)
        if file_extension in [".wav", ".flac"]:
            with open("test.wav", "wb") as f:
                f.write(audio_file.getbuffer())

            st.audio("test.wav", format="audio/wav")

        elif file_extension == ".mp3":
            with open("test.mp3", "wb") as f:
                f.write(audio_file.getbuffer())

            st.audio("test.mp3", format="audio/mpeg")

        #Load the WAV file using torchaudio
        if file_extension in [".wav", ".flac"]: 
            wav, sample_rate = torchaudio.load("test.wav")
            # st.markdown("Before unsquueze wav")
            # st.markdown(wav)
            file_extension_ori =".wav"
            #Unsqueeze for line 176
            wav= wav.unsqueeze(0)
                
        elif file_extension == ".mp3":
            # Load an MP3 file
            audio = AudioSegment.from_mp3("test.mp3")

            # Export it as a WAV file
            audio.export("test.wav", format="wav")
            wav3, sample_rate = torchaudio.load("test.wav")
            wav= wav3.unsqueeze(0)
            file_extension_ori =".mp3"
            file_extension =".wav"

            #RuntimeError: Could not infer dtype of numpy.float32
            #wav = torch.tensor(wav3).float() / 32768.0
            
            #RuntimeError: Numpy is not available
            # wav = torch.from_numpy(wav3) #/32768.0
            # wav = wav.unsqueeze(0).unsqueeze(0)
            # st.markdown("Before unsqueeze mp3")
            # st.markdown(wav)
        
        #Unsqueeze for line 176
        # wav= wav.unsqueeze(0)

        action = st.selectbox("Select Action", ["Add Watermark", "Detect Watermark"])

        if action == "Add Watermark":
            #watermark_text = st.text_input("The watermark (0, 1 list of length-16):", value=st.session_state.def_value)
            add_watermark_button = st.button("Add Watermark", key="add_watermark_btn")
            if add_watermark_button:  # 点击按钮后执行的
                #if audio_file and watermark_text:
                if audio_file:
                    with st.spinner("Adding Watermark..."):
                        #wav = my_read_file(wav,max_second_encode)
                        
                        #1st attempt
                        watermark = model.get_watermark(wav, default_sr)
                        watermarked_audio = wav + watermark
                        print(watermarked_audio.size())
                        size = watermarked_audio.size()
                        #st.markdown(size)

                        print(watermarked_audio.squeeze())
                        squeeze = watermarked_audio.squeeze(1)
                        shape = squeeze.size()
                        #st.markdown(shape)

                        #st.markdown(squeeze)
                    if file_extension_ori in [".wav", ".flac"]:
                        torchaudio.save("output.wav", squeeze, default_sr, bits_per_sample=16)
                        watermarked_wav = torchaudio.save("output.wav", squeeze, default_sr, bits_per_sample=16)

                        st.audio("output.wav", format="audio/wav")

                        with open("output.wav", "rb") as file:
                            #file.read()
                            #file.write(watermarked_wav.getbuffer())
                            binary_data = file.read()
                            btn = st.download_button(
                                label="Download watermarked audio",
                                data=binary_data,
                                file_name="output.wav",
                                mime="audio/wav",
                            )


                    elif file_extension_ori == ".mp3":
                        torchaudio.save("output.wav", squeeze, default_sr)
                        watermarked_mp3 = torchaudio.save("output.wav", squeeze, default_sr)
                        audio = AudioSegment.from_wav("output.wav")

                        # Export as MP3
                        audio.export("output.mp3", format="mp3")
                        st.audio("output.mp3", format="audio/mpeg")
                        
                        with open("output.mp3", "rb") as file:
                            #file.write(watermarked_wav.getbuffer())
                            binary_data = file.read()
                            st.download_button(
                                label="Download watermarked audio",
                                data=binary_data,
                                file_name="output.mp3",
                                mime="audio/mpeg",
                            )

        elif action == "Detect Watermark":
            detect_watermark_button = st.button("Detect Watermark", key="detect_watermark_btn")

            if audio_file:
                #1st attempt
                watermark = model.get_watermark(wav, default_sr)
                watermarked_audio = wav + watermark
                print(watermarked_audio.size())
                size = watermarked_audio.size()
                #st.markdown(size)

                
            if detect_watermark_button:
                with st.spinner("Detecting..."):
                    result, message = detector.detect_watermark(watermarked_audio, sample_rate=default_sr, message_threshold=0.5)
                    st.markdown("This is likely a watermarked audio:")
                    st.markdown(result)
                    print(f"\nThis is likely a watermarked audio: {result}")

                    # Run on an unwatermarked audio

                    if file_extension in [".wav", ".flac"]: 
                        wav, sample_rate = torchaudio.load("test.wav")
                        wav= wav.unsqueeze(0)

                    elif file_extension == ".mp3":
                        # Load an MP3 file
                        audio = AudioSegment.from_mp3("test.mp3")
                        # Export it as a WAV file
                        audio.export("test.wav", format="wav")
                        wav, sample_rate = torchaudio.load("test.wav")
                        wav= wav.unsqueeze(0)
                        
                    result2, message2 = detector.detect_watermark(wav, sample_rate=default_sr, message_threshold=0.5)
                    st.markdown("This is likely an unwatermarked audio:")
                    print(f"This is likely an unwatermarked audio: {result2}")
                    st.markdown(result2)


if __name__ == "__main__":
    default_sr = 16000
    max_second_encode = 60
    max_second_decode = 30
    len_start_bit = 16
    device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
    # model = wavmark.load_model().to(device)
    model = AudioSeal.load_generator("audioseal_wm_16bits")
    detector = AudioSeal.load_detector(("audioseal_detector_16bits"))
    main()

    # audio_path = "/Users/my/Library/Mobile Documents/com~apple~CloudDocs/CODE/PycharmProjects/4_语音水印/419_huggingface水印/WavMark/example.wav"

    # decoded_watermark, decode_cost = decode_watermark(audio_path)
    # print(decoded_watermark)