# Import necessary libraries import streamlit as st from pydub import AudioSegment import speech_recognition as sr import wikipedia import openai from gtts import gTTS import tempfile import os # Set OpenAI API key (replace with your actual API key) openai.api_key = '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' # Function to convert the uploaded audio file to WAV format if it's not already in WAV format def convert_to_wav(audio_file): file_extension = audio_file.name.split('.')[-1].lower() if file_extension == "mp3": audio = AudioSegment.from_mp3(audio_file) wav_file = tempfile.NamedTemporaryFile(delete=False, suffix=".wav") audio.export(wav_file.name, format="wav") return wav_file.name elif file_extension in ["wav", "flac", "aiff"]: return audio_file else: raise ValueError("Unsupported audio format. Please upload MP3, WAV, or FLAC.") # Function to recognize speech from the audio file def recognize_speech_from_file(file_path): recognizer = sr.Recognizer() # Open the audio file and recognize the speech with sr.AudioFile(file_path) as source: audio = recognizer.record(source) # Read the entire audio file try: print("Recognizing...") text = recognizer.recognize_google(audio) # Use Google's speech recognition return text except sr.UnknownValueError: return "Sorry, I could not understand the audio." except sr.RequestError: return "Sorry, my speech service is down." # Function to get the Wikipedia summary def get_wikipedia_answer(query): try: result = wikipedia.summary(query, sentences=2) # Get a short summary from Wikipedia return result except wikipedia.exceptions.DisambiguationError as e: return f"Your query is ambiguous, did you mean: {e.options}" except wikipedia.exceptions.HTTPTimeoutError: return "I am unable to access Wikipedia right now." except wikipedia.exceptions.RedirectError: return "Redirect error while fetching the information." except wikipedia.exceptions.PageError: return "Sorry, I couldn't find any information on that topic." # Function to get the answer from GPT-3/4 (OpenAI) def get_gpt_answer(query): try: response = openai.Completion.create( engine="text-davinci-003", # Use the appropriate engine based on your OpenAI account prompt=query, max_tokens=150 ) return response.choices[0].text.strip() except openai.error.OpenAIError as e: return f"Error from GPT: {str(e)}" # Function to convert text to speech using gTTS def speak(text): tts = gTTS(text=text, lang='en') tts.save("response.mp3") # Save the speech to a file return "response.mp3" # Streamlit UI def main(): st.title("Voice-to-Text Chatbot") st.write("Upload an audio file to ask a question, and the bot will respond with text and speech.") # Upload audio file uploaded_file = st.file_uploader("Upload an Audio File", type=["mp3", "wav", "flac"]) if uploaded_file is not None: # Convert the uploaded file to WAV if necessary audio_file_path = convert_to_wav(uploaded_file) st.audio(uploaded_file, format='audio/wav') # Play the uploaded audio for the user # Recognize speech from the audio file user_input = recognize_speech_from_file(audio_file_path) st.write(f"You asked: {user_input}") # Get the answer from Wikipedia or GPT-3 wikipedia_answer = get_wikipedia_answer(user_input) if wikipedia_answer: st.write(f"Wikipedia says: {wikipedia_answer}") response_text = wikipedia_answer else: gpt_answer = get_gpt_answer(user_input) st.write(f"GPT says: {gpt_answer}") response_text = gpt_answer # Convert the response to speech and play it response_audio_path = speak(response_text) st.audio(response_audio_path, format='audio/mp3') # Play the response audio if __name__ == "__main__": main()