Mengenali ucapan menggunakan IBM Speech-to-Text API

Hai, Habr. Sebagai bagian dari kursus "Machine Learning. Advanced", kami telah menyiapkan terjemahan materi yang menarik untuk Anda.



Kami juga mengundang semua orang untuk menonton pelajaran terbuka tentang topik "Multi-strategi untuk mengoptimalkan pengujian AB".

---

, Python.

, IBM Speech to Text API . API, , , , . API . .

, , « » . , , — .

— , , - . . - . , , . .

. . , .

:

• 
  
  
  
  

• 1 -

  
  
  
  
  

• 2 -

  
  
  
  
  

• 3 -

  
  
  
  
  

• 4 -

  
  
  
  
  

• -

  
  
  
  
  

. , . API- , . , , . , Microsoft Azure.

:

• Google Cloud

  
  
  
  
  

• IBM Watson

  
  
  
  
  

• Microsoft Azure

  
  
  
  
  

• Amazon Transcribe

  
  
  
  
  

1 —

. SpeechRecognition



. SpeechRecognition



. API. ; Microsoft Azure Speech, Google Cloud Speech, API IBM Watson Speech to Text . IBM Watson Speech to Text API. SpeechRecognition .

. pip, Python.

pip install SpeechRecognition
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

. Jupyter Notebook.

import speech_recognition as s_r
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

2 -

. m4a, m4a. wav .

audio_file = s_r.AudioFile('my_clip.wav')
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

3 -

, , . . .

rcgnzr = s_r.Recognizer()
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

4 -

! IBM speech to text . , “adjust_for_ambient_noise



” “record



”, . , .

with audio_file as source: 
   rcgnzr.adjust_for_ambient_noise(source) 
   clean_audio = rcgnzr.record(source)
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

, . IBM. ( , , IBM Speech-to-Text API Python SpeechRecogniton



). API:

recognized_speech_ibm = r.recognize_ibm(clean_audio, username="apkikey", password= "your API Key")
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

: API IBM API-. IBM Watson. Speech-to-Text . IBM, , 500 , , .

-

. . . , . , .

, :

print(recognized_speech_ibm)
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

. . . , . , ​​ , .

. «ready!» .

with open('recognized_speech.txt',mode ='w') as file:    
   file.write("Recognized Speech:") 
   file.write("\n") 
   file.write(recognized) 
   print("ready!")
      
      

        
        
        
      

    
        
        
        
      
      

        
        
        
      

    
    

! , . , - . — . , . , .

. YouTube, . .

---

"Machine Learning. Advanced".

«Multi-armed bandits AB ».