app.py

import streamlit as st
import speech_recognition as sr
from deep_translator import GoogleTranslator
from gtts import gTTS
from playsound import playsound
from PIL import Image
from streamlit_mic_recorder import mic_recorder,speech_to_text
from transformers import AutoTokenizer, M2M100ForConditionalGeneration
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
from transformers import Wav2Vec2ForCTC, Wav2Vec2Tokenizer
from sentence_transformers import SentenceTransformer, util
import torch
from pygame import mixer
from io import BytesIO
import httpcore
setattr(httpcore, 'SyncHTTPTransport', 'AsyncHTTPProxy')
# pygame.init()
# import logging
# import logging.handlers
# import queue
# import threading
# import time
# import urllib.request
# import os
# from collections import deque
# from pathlib import Path
# from typing import List

# import av
# import numpy as np
# import pydub
# import streamlit as st
# from twilio.rest import Client

# from streamlit_webrtc import WebRtcMode, webrtc_streamer
#from streamlit_mic_recorder import mic_recorder,speech_to_text
# Initialize the recognizer
recognizer = sr.Recognizer()

# Initialize the translator


# Language dictionary
language_dict = {
    'bengali': 'bn',
    'english': 'en',
    'gujarati': 'gu',
    'hindi': 'hi',
    'malayalam': 'ml',
    'marathi': 'mr',
    'nepali': 'ne',
    'odia': 'or',
    'punjabi': 'pa',
    'tamil': 'ta',
    'telugu': 'te',
}


# Function to translate speech
def translate_speech():
    
    #source_language_name = recognize_speech("Please speak the source language name (e.g., 'English'): ")
    
        
    st.title("BITranSlate")
    # st.write("Record your voice, and play the recorded audio:")
    # audio=mic_recorder(start_prompt="⏺️",stop_prompt="⏹️",key='recorder')

    custom_theme = {
            "theme": {
                "primaryColor": "#000000",
                "backgroundColor": "#89939E",
                "secondaryBackgroundColor": "#262730",
                "textColor": "#FFFFFF",
                "font": "Serif"
            }
        }
    st.markdown(
            f"""
            <style>
            :root {{
                --primary-color: {custom_theme["theme"]["primaryColor"]};
                --background-color: {custom_theme["theme"]["backgroundColor"]};
                --secondary-background-color: {custom_theme["theme"]["secondaryBackgroundColor"]};
                --text-color: {custom_theme["theme"]["textColor"]};
                --font: {custom_theme["theme"]["font"]};
            }}
            </style>
            """,
            unsafe_allow_html=True
        )
    source_language_name = st.selectbox('Please input the source language',language_dict)
    source_language = language_dict[source_language_name]
    target_language_name = st.selectbox('Please input the target language',language_dict)
    target_language = language_dict[target_language_name]

    c1,c2=st.columns(2)
    with c1:
        st.write("Convert speech to text:")
    with c2:
        text=speech_to_text(language=source_language,use_container_width=True,just_once=True,key='STT')

    sentence = text
    nllb_langs = {'hindi':'hin_Deva',
                  'english':'eng_Latn',
                  'punjabi':'pan_Guru',
                  'odia':'ory_Orya',
                  'bengali':'ben_Beng',
                  'telugu':'tel_Tulu',
                  'tamil':'tam_Taml',
                  'nepali':'npi_Deva',
                  'marathi':'mar_Deva',
                  'malayalam':'mal_Mlym',
                  'kannada':'kan_Knda',
                  'gujarati':'guj_Gujr',
                  }
    # translator_google = Translator(service_urls=[
    #   'translate.googleapis.com'
    # ])
    #translator_google = google_translator()
    translator = pipeline('translation', model=AutoModelForSeq2SeqLM.from_pretrained("facebook/nllb-200-distilled-600M"), tokenizer=AutoTokenizer.from_pretrained("facebook/nllb-200-distilled-600M"), src_lang=nllb_langs[source_language_name], tgt_lang=nllb_langs[target_language_name], max_length = 4000)
    text_to_translate = text
    translated_text = translator(text_to_translate)[0]['translation_text']
    translated_text_google = GoogleTranslator(source='auto', target=target_language).translate(text_to_translate)
    #translated_text_google = translator_google.translate(text_to_translate, lang_tgt=target_language)

    # translated_text_google = translator_google.translate(text_to_translate, src=source_language, dest=target_language)

    #translated_text_google = translator_google.translate(text_to_translate, src=source_language, dest=target_language)
    model2 = SentenceTransformer("google/muril-base-cased")
            # Compute embeddings for the sentences
    embedding = model2.encode(text_to_translate, convert_to_tensor=True)
    embeddings_nllb = model2.encode(translated_text, convert_to_tensor=True)
    embeddings_google = model2.encode(translated_text_google, convert_to_tensor=True)

    # Calculate cosine similarities
    cosine_score_nllb = util.cos_sim(embedding, embeddings_nllb).item()
    cosine_score_google = util.cos_sim(embedding, embeddings_google).item()

    # Select the translation with the higher cosine similarity score
    selected_translation = translated_text if cosine_score_nllb > cosine_score_google else translated_text_google

    st.write(f"Source Language: {source_language_name}")
    st.write(f"Sentence: {sentence}")
    st.write(f"Destination Language: {target_language_name}")
    st.write(f"Translated Text from NLLB: {translated_text}")
    st.write(f"Translated Text from Google Translate: {translated_text_google}")
    st.write(f"More accurate translation: {selected_translation}")

    # Using Google-Text-to-Speech to speak the translated text
    # speak = gTTS(text=translated_text, lang=target_language, slow=False)
    # speak.save("translated_voice.mp3")
    mp3_fp = BytesIO()
    speak = gTTS(text=translated_text, lang=target_language, slow=False)
    speak.write_to_fp(mp3_fp)

    # Play the translated voice
    mixer.init()
    mp3_fp.seek(0)
    mixer.music.load(mp3_fp, "mp3")
    mixer.music.play()
    # Play the translated voice
    # mixer.init()
    # mixer.music.load('translated_voice.mp3')
    # mixer.music.play()
    #playsound('translated_voice.mp3')

#if st.button("  CLICK HERE TO TRANSLATE  "):
translate_speech()