app.py

import gradio as gr
import speech_recognition as sr
from io import BytesIO
from pydub import AudioSegment

def transcribe_audio(audio_input):
    recognizer = sr.Recognizer()
    
    # Ensure the input is a tuple and get the audio data bytes
    if isinstance(audio_input, tuple) and len(audio_input) == 2:
        audio_data_bytes = audio_input[1]
    else:
        raise ValueError("Expected audio_input to be a tuple with audio data bytes.")

    # Use BytesIO to create a file-like object from the audio bytes
    audio_file = BytesIO(audio_data_bytes)
    
    # Convert audio to WAV format using pydub
    audio_segment = AudioSegment.from_file(audio_file)
    wav_io = BytesIO()
    audio_segment.export(wav_io, format="wav")
    wav_io.seek(0)  # Move to the beginning of the file-like object
    
    # Load the audio file from the file-like object in WAV format
    with sr.AudioFile(wav_io) as source:
        audio_data = recognizer.record(source)
        
    try:
        # Transcribe the audio data
        text = recognizer.recognize_google(audio_data)
        return text
    except sr.UnknownValueError:
        return "Google Speech Recognition could not understand audio"
    except sr.RequestError as e:
        return f"Could not request results from Google Speech Recognition service; {e}"

# Create the Gradio interface
iface = gr.Interface(
    fn=transcribe_audio,
    inputs="audio",
    outputs="text",
    title="Voice to Text Converter",
    description="Upload an audio file and get the transcribed text."
)

# Launch the interface
iface.launch()