VietGPT-VoiceBot-A-Vietnamese-Speech-Recognition-Chatbot

VietGPT VoiceBot: Chatbot automatically recognizes Vietnamese voice and uses the ChatGPT API for natural language interaction.

Installation

PyAudio:

Use pip to install PyAudio:
```
pip install pyaudio
```

Requests:

Use pip to install Requests:
```
pip install requests
```

gTTS (Google Text-to-Speech):

Use pip to install gTTS:
```
pip install gtts
```

Keyboard:

Use pip to install Keyboard:
```
pip install keyboard
```

Pygame:

Use pip to install pygame
```
pip install pygame
```

Transformers:

Use pip to install Transformers
```
pip install transformers
```

Record audio

I use python's pyaudio library to record voice and write it to an input .wav file

record_audio

def record_audio(duration=5):
CHUNK = 1024
FORMAT = pyaudio.paInt16
CHANNELS = 1
RATE = 16000

p = pyaudio.PyAudio()

stream = p.open(format=FORMAT,
                channels=CHANNELS,
                rate=RATE,
                input=True,
                frames_per_buffer=CHUNK)

print("Hold Space bar to start recording...\n")

frames = []
is_recording = False
while True:
    if keyboard.is_pressed('space'):
        if not is_recording:
            mixer.quit()
            is_recording = True
    else:
        if is_recording:
            time.sleep(2)
            break

    if is_recording:
        data = stream.read(CHUNK)
        frames.append(data)

stream.stop_stream()
stream.close()
p.terminate()

wf = wave.open(filename, 'wb')
wf.setnchannels(CHANNELS)
wf.setsampwidth(p.get_sample_size(FORMAT))
wf.setframerate(RATE)
wf.writeframes(b''.join(frames))
wf.close()

Speech auto Recognition

I use Phowhisper's transformers library developed by VinAI Company to listen and understand Vietnamese or English language, from which I can output the words you say.

Transcrible language

  transcriber = pipeline("automatic-speech-recognition", model="vinai/PhoWhisper-small")
  output = transcriber(filename)['text']

Connect openAI

I use the API provided by openAI to chat, interact and ask questions with chatGPT.

Connect with openAI

  def GetResultFromOpenAI(text):
  openai_api_key = "openai_api_key"
  url = "https://api.openai.com/v1/chat/completions"
      
  headers = {
      "Content-Type": "application/json",
      "Authorization": f"Bearer {openai_api_key}"
  }

  data = {
      "model": "gpt-3.5-turbo",
      "messages": json_data_list,
      "temperature": 1.1
  }
  response = requests.post(url, headers=headers, json=data)

  if response.status_code == 200:
      result = response.json()['choices'][0]['message']['content']
      return result
  else:
      print("Error:", response.status_code, response.text)

Text to speech with Google

I use the gTTS library made by Google to convert text to audio and play it to listener

Text to speech

def text_to_speech_vietnamese(text):
mixer.quit()
if os.path.exists(output_file):
    os.remove(output_file)
    
tts = gTTS(text=text, lang='vi')
tts.save(output_file)

mixer.init()
mixer.music.load(output_file)
mixer.music.play()

Authors

BUI QUANG VINH
Contact: BUI QUANG VINH

Donations

Support the ongoing development and improvement of SeleniumSupport:

Donate via Binance Address: TRvXSzxnjDWTTi1fnouqjFvb2YNzG5RqsZ
Donate via Momo Address: 0974602103
Donate via Paypal Address: quangvinhb167@gmail.com
Support on Telegram
Support on Facebook

bemka1/VietGPT-VoiceBot-A-Vietnamese-Speech-Recognition-Chatbot