app.py

import streamlit as st
import speech_recognition as sr
from deep_translator import GoogleTranslator
from pydub import AudioSegment
from io import BytesIO
import tempfile

# Title of the app
st.title("Speech-to-Text with Translation to English")

# Initialize recognizer
recognizer = sr.Recognizer()

# Choice for input language
language_options = {"English": "en", "Hindi": "hi"}
input_language = st.selectbox("Select Input Language", options=language_options.keys())
selected_lang_code = language_options[input_language]

# Function to convert audio chunk to text
def speech_to_text(audio_data, lang="en"):
    try:
        st.info("Converting speech to text...")
        detected_text = recognizer.recognize_google(audio_data, language=lang)
        return detected_text
    except Exception as e:
        st.error(f"Error in speech recognition: {e}")
        return None

# Process uploaded audio file
uploaded_file = st.file_uploader("Upload an audio file", type=["wav", "mp3", "ogg"])
if uploaded_file:
    with st.spinner("Processing uploaded audio..."):
        try:
            # Convert uploaded file to WAV format using pydub
            audio = AudioSegment.from_file(BytesIO(uploaded_file.read()))
            # Split audio into 30-second chunks
            chunk_duration_ms = 30000
            chunks = [audio[i:i+chunk_duration_ms] for i in range(0, len(audio), chunk_duration_ms)]
            text_output = ""

            for i, chunk in enumerate(chunks):
                with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_wav_file:
                    chunk.export(tmp_wav_file.name, format="wav")
                    with sr.AudioFile(tmp_wav_file.name) as source:
                        audio_data = recognizer.record(source)
                        detected_text = speech_to_text(audio_data, lang=selected_lang_code)
                        if detected_text:
                            text_output += detected_text + " "

            # Display detected text and translate
            if text_output:
                st.write("Detected Speech Text:", text_output)
                translator = GoogleTranslator(source='auto', target='en')
                translated_text = translator.translate(text_output)
                st.write("Translated Text (English):", translated_text)

        except Exception as e:
            st.error(f"Error processing the audio file: {e}")