pydub、playsound、pygame播放声音；gradio、streamlit页面播放声音；gradio 页面图像、视频及调用摄像头_gradio调用摄像头

1、pydub

from pydub import AudioSegment
from pydub.playback import play

song = AudioSegment.from_wav(r"C:\Users\loong\Downloads\zh.wav")
play(song)
1
2
3
4
5

2、playsound

from playsound import playsound

playsound(r"voice.wav")
1
2
3

报错：
参考：https://blog.csdn.net/weixin_50836014/article/details/122135430
https://blog.csdn.net/lj606/article/details/122354958
https://stackoverflow.com/questions/68826091/the-specified-device-is-not-open-or-is-not-recognized-by-mci （这个说降版本，绝对路径）

raise PlaysoundException(exceptionMessage)
playsound.PlaysoundException:
Error 263 for command:
open audio123.wav
指定的设备未打开，或不被 MCI 所识别。

解决方法：去C:\Use***\Python310\site-packages\playsound.py 下更改去注释掉uft-16；两个地方修改

报错：ERROR:playsound:
Error 296 for command:
open “D:\llm\edgetts1.wav”
无法在指定的 MCI 设备上播放指定的文件。文件可能已损坏，或格式不对，或没有此格式的文件处理程序可用。
ERROR:playsound:
Error 263 for command:
close “D:\llm\edgetts1.wav”
指定的设备未打开，或不被 MCI 所识别。
WARNING:playsound:Failed to close the file: “D:\llm\edgetts1.wav”

这里是edge tts生成语音文件用playsound来播放；生成如果是wav格式就报错，生成mp3文件就可以正常播放
参考：https://blog.csdn.net/weixin_44493841/article/details/134531993

pygame

import pygame

def my_playsound(filePath):
    
    pygame.init()
    track = pygame.mixer.music.load(filePath)
 
    # 播放
    pygame.mixer.music.play()
    
    # 等待直到播放结束
    while pygame.mixer.music.get_busy():
        for event in pygame.event.get():
            if event.type == pygame.QUIT:
                pygame.mixer.music.stop()
                sys.exit()
    pygame.quit()
    
my_playsound(OUTPUT_FILE)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19

3、streamlit

import streamlit as st
##输入字节流
st.audio(audio_bytes) ##展示前端语音控件
1
2
3

离线音频自动播放

import base64

import streamlit as st


## 文件
file_path = "audio123.wav"

with open(file_path, "rb") as f:
        audio_bytes = f.read()

        audio_base64 = base64.b64encode(audio_bytes).decode('utf-8')
        audio_tag = f'<audio autoplay="true" src="data:audio/wav;base64,{audio_base64}">'
        st.markdown(audio_tag, unsafe_allow_html=True)
1
2
3
4
5
6
7
8
9
10
11
12
13
14

1）录制声音

2）录制声音 audio-recorder-streamlit （这款很棒一个点是可以设置pause_threshold，没有说话声音多久就自动停止录制）

https://github.com/Joooohan/audio-recorder-streamlit
安装：

pip install audio-recorder-streamlit
1

使用：

import streamlit as st
from audio_recorder_streamlit import audio_recorder

audio_bytes = audio_recorder(pause_threshold=1.5)
if audio_bytes:
    st.audio(audio_bytes, format="audio/wav")
1
2
3
4
5
6

录制保存wav

import streamlit as st
from audio_recorder_streamlit import audio_recorder


def record():
    audio_bytes = audio_recorder(text="点击说话", pause_threshold=1.5)
    if audio_bytes:
        st.audio(audio_bytes)
        with open("streamlit_record.wav", mode='bw') as f:
            f.write(audio_bytes)
            f.close()
    return audio_bytes


audio_data = record()
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

https://github.com/bouzidanas/streamlit-float 另外streamlit-float 可以控制让麦克风图标一直保持在页面具体位置，不随着页面下拉变化

import streamlit as st
from audio_recorder_streamlit import audio_recorder

from streamlit_float import *

# Float feature initialization
float_init()

footer_container = st.container()

def record():
    # Create footer container for the microphone
    with footer_container:
        audio_bytes = audio_recorder(text="点击说话",pause_threshold=1.5)
    if audio_bytes:
        st.audio(audio_bytes)
        with open("streamlit_record.wav", mode='bw') as f:
            f.write(audio_bytes)
            f.close()
    return audio_bytes

audio_data = record()
# Float the footer container and provide CSS to target it with
footer_container.float("bottom: 0rem;")







1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

4、gradio

版本：gradio-4.16.0 在3.38不生效

gr.Audio既可以输入也可以播放这个函数接口
inputs=gr.Audio(type=“file”, label=“上传音频文件”), # 输入组件，允许用户上传音频文件
outputs=gr.Audio(label=“播放音频”, autoplay=True)

import gradio as gr

def greet(name):
    return "Hello " + name + "!"

# 创建 Gradio 界面
with gr.Blocks() as demo:
    with gr.Row():
        inp_ref = gr.Audio(label="请上传3~10秒内参考音频，超过会报错！", sources=["microphone", "upload"], type="filepath")
        gr.Markdown("### Greet")
        # gr.Input(type="text", label="Name:")
    with gr.Row():
        gr.Button("Greet")
        gr.Markdown("### Response")
        # gr.Markdown("### ", output=True)

# 启动应用
demo.launch()
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18

注意：传入 type="filepath"才会有文件路径值

5、gradio 页面图像、视频及调用摄像头

gr.Image(
type=‘numpy’,
label=‘Input Image’
)
gr.Video(
label=‘Input Video’
)

with gr.Row():
        input_image_component = gr.Image(
                    type='numpy',
                    label='Input Image'
                )
        output_image_component = gr.Image(
            type='numpy',
            label='Output Image'
                )
        
    with gr.Row():
        input_video_component = gr.Video(
            label='Input Video'
        )
        output_video_component = gr.Video(
            label='Output Video'
        )
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17