首先新建AipSpeech（百度语音识别的SDK客户端）
""" 百度AI开放平台的 APPID AK SK """
APP_ID = '你的 App ID'
API_KEY = '你的 Api Key'
SECRET_KEY = '你的 Secret Key'
client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
然后定义语音到文本的功能。
[En]
Then define a function of voice to text.
path = 'voices/voice.wav'
def listen():
with open(path, 'rb') as fp:
voice = fp.read()
result = client.asr(voice, 'wav', 16000, {'dev_pid': 1537})
try:
result_text = result["result"][0]
print(result_text)
except KeyError:
print("KeyError")
print(result)
定义录音的函数
def record_audio(rate=16000):
r = sr.Recognizer()
with sr.Microphone(sample_rate=rate) as source:
print("please say something")
r.adjust_for_ambient_noise(source)
audio = r.listen(source,phrase_time_limit=59)
print('record finish')
with open("voices/voice.wav", "wb") as f:
f.write(audio.get_wav_data())
Microphone()的使用方法
实例化Recognize()后调用下面的方法（这里面的r指的就是下面的recognizer_instence)
用speech_recognition模块可以实现有静音识别的录音（还可以与snowboy集成，实现热词识别）
动态调整能量门限，解决环境噪声(自动调整静音检测门限)(使用后效果明显提升)
[En]
Dynamically adjust the energy threshold to solve the ambient noise (automatically adjust the threshold of mute detection) (the effect can be significantly improved after use)
参考手册：https://github.com/Uberi/speech_recognition/blob/master/reference/library-reference.rst
完整代码：
import speech_recognition as sr
from aip import AipSpeech
APP_ID = '18490357'
API_KEY = 'PdZpve6uqC7qg8kgW7LsazWZ'
SECRET_KEY = '4fNWQFkTeyxAku412byYYoRArBDGkiSg'
client = AipSpeech(APP_ID, API_KEY, SECRET_KEY)
path = 'voices/voice.wav'
def record_audio(rate=16000):
r = sr.Recognizer()
with sr.Microphone(sample_rate=rate) as source:
print("please say something")
r.adjust_for_ambient_noise(source)
audio = r.listen(source,phrase_time_limit=59)
print('record finish')
with open("voices/voice.wav", "wb") as f:
f.write(audio.get_wav_data())
def listen():
with open(path, 'rb') as fp:
voice = fp.read()
result = client.asr(voice, 'wav', 16000, {'dev_pid': 1537})
try:
result_text = result["result"][0]
print(result_text)
except KeyError:
print("KeyError")
print(result)
if __name__ == "__main__":
while True:
record_audio()
listen()
Original: https://blog.csdn.net/milk_paramecium/article/details/110094565
Author: 喝牛奶的草履虫
Title: 实现自动连续的语音转文字，使用speech_recognition实现静音检测的录音，并使用百度AI开放平台的语音转文字接口将语音转文本
原创文章受到原创版权保护。转载请注明出处：https://www.johngo689.com/526565/
转载文章受原作者版权保护。转载请注明原作者出处！