python speech to text

1The best library because you dont have to save the
2text file or open the file to start the speech
3
4pip install pyttsx3
5
6import pyttsx3
7engine = pyttsx3.init()
8engine.say("Hello world")
9engine.runAndWait()

1#pip install SpeechRecognition
2#in case of error use 'pip install pyaudio' or...
3#in case of error use 'pip install pipwin' then 'pipwin install pyaudio'
4#if error continued you may need to use python 3.6 or lower as the latest 
5#python may not support pyaudio... 
6import speech_recognition as sr
7import pyttsx3
8
9#audio of system to respond
10engine = pyttsx3.init('sapi5')
11voices = engine.getProperty('voices')
12engine.setProperty('voice', voices[0].id)
13engine.setProperty('rate',180)
14
15def speak(audio):
16    engine.say(audio)
17    engine.runAndWait()
18
19# simple function to recognise speech from user
20def takecommand():
21    #it takes microphone input and returns string output
22    r = sr.Recognizer()
23    with sr.Microphone() as source:
24        print('Listening.....')
25        r.pause_threshold = 1
26        r.energy_threshold = 4000
27        audio = r.listen(source)
28
29    try:
30        print('Recognising...')
31        query = r.recognize_google(audio, language='en-in')
32        print('User Said : ' , query)
33
34    except Exception as e:
35        print('exception : ',e)
36
37        speak("Sorry, I didn't hear that, Say that again Please")
38        return "None"
39    return query
40while True:
41  query = takecommand() # whatever user says will be stored in this variable
42  print("The Test got in program is : "+query)
43

1import speech_recognition as sr
2
3def take_command():
4    r = sr.Recognizer()
5    with sr.Microphone() as source:
6        print('Listening...')
7        r.pause_threshold = 1
8        r.energy_threshold = 50
9        audio = r.listen(source)
10
11    try:
12        print('Recognizing...')
13        qry = r.recognize_google(audio, language='en-in')
14        print(f"user said: {qry}\n")
15        
16#     if any error occurs this line will run
17    except Exeption as e:
18    # if you don't want to print the error comment the bottom line
19        print(e)
20        print('Say that again please\n')
21        return 'None'
22
23    return qry
24  
25if __name__ == '__main__':
26	while True:
27  		qry = takecommand().lower()
28  
29# now you can use the takecommand function where you want to recognize speech
30# And please experiment with the above code 
31# like what pause_threshold and energy_threshold do 
32/\/\/\/\/\/\/\/\/\/\/\---  *HAPPYCODING*  ---/\/\/\/\/\/\/\/\/\/\/\

1  import speech_recognition as sr
2
3
4        def main():
5
6            r = sr.Recognizer()
7
8            with sr.Microphone() as source:
9                r.adjust_for_ambient_noise(source)
10
11                audio = r.listen(source)
12
13                try:
14
15                    print(r.recognize_google(audio))
16
17                except Exception as e:
18                    print("Error :  " + str(e))
19
20
21                with open("recorded.wav", "wb") as f:
22                    f.write(audio.get_wav_data())
23
24
25        if __name__ == "__main__":
26            main()

1# importing libraries 
2import speech_recognition as sr 
3import os 
4from pydub import AudioSegment
5from pydub.silence import split_on_silence
6
7# create a speech recognition object
8r = sr.Recognizer()
9
10# a function that splits the audio file into chunks
11# and applies speech recognition
12def get_large_audio_transcription(path):
13    """
14    Splitting the large audio file into chunks
15    and apply speech recognition on each of these chunks
16    """
17    # open the audio file using pydub
18    sound = AudioSegment.from_wav(path)  
19    # split audio sound where silence is 700 miliseconds or more and get chunks
20    chunks = split_on_silence(sound,
21        # experiment with this value for your target audio file
22        min_silence_len = 500,
23        # adjust this per requirement
24        silence_thresh = sound.dBFS-14,
25        # keep the silence for 1 second, adjustable as well
26        keep_silence=500,
27    )
28    folder_name = "audio-chunks"
29    # create a directory to store the audio chunks
30    if not os.path.isdir(folder_name):
31        os.mkdir(folder_name)
32    whole_text = ""
33    # process each chunk 
34    for i, audio_chunk in enumerate(chunks, start=1):
35        # export audio chunk and save it in
36        # the `folder_name` directory.
37        chunk_filename = os.path.join(folder_name, f"chunk{i}.wav")
38        audio_chunk.export(chunk_filename, format="wav")
39        # recognize the chunk
40        with sr.AudioFile(chunk_filename) as source:
41            audio_listened = r.record(source)
42            # try converting it to text
43            try:
44                text = r.recognize_google(audio_listened)
45            except sr.UnknownValueError as e:
46                print("Error:", str(e))
47            else:
48                text = f"{text.capitalize()}. "
49                print(chunk_filename, ":", text)
50                whole_text += text
51    # return the text for all chunks detected
52    return whole_text

1pip install pyttsx3
2