I. Overview
This article briefly introduces the basic use of the voice recognition of Baidu (in fact, when the landlord wants to get a card player and no money, grab a bag of what is not, had to make speech recognition)
Second, create the application
Open the Baidu Voice official website, product and use --- voice recognition--now use --Create an app
The following page appears
Follow the prompts to complete the final result
(PS: I just want to get a card, it's the name of a counter)
Click "View Key" on the right to note the app Id,api Key,secret key. The next step is to use
need to install module PIP install Baidu-aippip install Pyaudio
Speech Recognition Code
fromAIP Import Aipspeech"""your APPID AK SK"""app_id='the app_id you wrote down.'Api_key='the Api_key you wrote down.'Secret_key='the Secret_key you wrote down.'Client=Aipspeech (app_id, Api_key, Secret_key) # Read file Def get_file_content (FilePath): With open (FilePath,'RB') asfp:returnFp.read () # identify local files Li=client.asr (Get_file_content ('01.PCM'),'PCM',8000, { 'LAN':'ZH',}) Print (LI) # get file recognition # from URL # client.asr ("','PCM',16000, {# 'URL':'HTTP://121.40.195.233/RES/16K_TEST.PCM',# 'Callback':'http://xxx.com/receive',# })
Python Recording Code
Import Wave fromPyaudio Import Pyaudio,paint16framerate=8000Num_samples= -Channels=1Sampwidth=2 Time=2def save_wave_file (filename,data):" "Save the date to the Wavfile" "WF=wave.open (filename,'WB') wf.setnchannels (Channels) wf.setsampwidth (sampwidth) wf.setframerate (framerate) wf.writeframes (b"". Join (data)) Wf.close () def my_record (): PA=Pyaudio () stream=pa.open (format = paint16,channels=1, rate=framerate,input=True, Frames_per_buffer=num_samples) My_buf=[] Count=0 whilecount<time*5: #控制录音时间 string_audio_data=Stream.read (num_samples) my_buf.append (string_audio_data) Count+=1Print ('.') Save_wave_file ('01.PCM', My_buf) stream.close () chunk= thedef play (): WF=wave.open (R"01.PCM",'RB') P=Pyaudio () stream=p.open (Format=p.get_format_from_width (Wf.getsampwidth ()), channels=wf.getnchannels (), rate=wf.getframerate (), output=True) whileTrue:data=wf.readframes (Chunk)ifdata=="": Breakstream.write (data) Stream.Close () p.terminate ()if__name__ = ='__main__': My_record () print ('over!') play ()
Effects such as:
Help Documentation:
Baidu Voice help document or manual
Third, PostScript
This code is not fully implemented, interested in self-organizing, playing landlords when the sound may be larger, because the recognition sometimes reported 3001 error, audio quality is too poor, but be beaten can not find me
Python calls Baidu Voice (speech recognition-bucket landlord voice card)