Upload 6 files

main.py

@@ -0,0 +1,33 @@
+import streamlit as st
+from  speechtotext import speech_to_text
+from t2s import text_to_speech_page
+
+def main():
+    st.title("VOICE TRANSFORMATION TECHNOLOGIES : EXPLORATION")
+
+    # Define a dictionary to map page names to their corresponding functions
+    pages = {
+        "Text to Speech": text_to_speech_page,
+        "Speech to Text": speech_to_text
+    }
+
+    # Display a sidebar for navigation
+    st.sidebar.title("Navigation")
+
+    # Get the selection from the user
+    selection = st.sidebar.radio("Go to", list(pages.keys()))
+
+    # If the user selects "Text to Speech", show the text-to-speech page
+    if selection == "Text to Speech":
+        text_to_speech_page()
+
+    # If the user selects "Speech to Text", show the speech-to-text page
+    elif selection == "Speech to Text":
+        speech_to_text()
+
+    # Add additional pages here as needed
+    # elif selection == "Page Name":
+    #     page_function()
+
+if __name__ == "__main__":
+    main()

s2s.py

@@ -0,0 +1,41 @@
+import speech_recognition as sr
+import pyttsx3
+
+# Function to transcribe audio to text
+def transcribe_audio(audio_file):
+    recognizer = sr.Recognizer()
+    with sr.AudioFile(audio_file) as source:
+        audio_data = recognizer.record(source)
+        try:
+            text = recognizer.recognize_sphinx(audio_data)
+            return text
+        except sr.UnknownValueError:
+            return "Speech recognition could not understand audio"
+        except sr.RequestError as e:
+            return "Error with speech recognition service: {0}".format(e)
+
+# Function to generate audio from text
+def generate_audio(text, output_file, rate=150):
+    engine = pyttsx3.init()
+    engine.setProperty('rate', rate)
+    engine.save_to_file(text, output_file)
+    engine.runAndWait()
+
+# Main code
+if __name__ == "__main__":
+    # Example audio file (replace with your audio file)
+    audio_file = "input_audio.wav"
+
+    # Transcribe audio to text
+    transcribed_text = transcribe_audio(audio_file)
+
+    if transcribed_text:
+        print("Transcribed text:", transcribed_text)
+
+        # Generate audio from transcribed text
+        output_file = "output_audio.wav"
+        generate_audio(transcribed_text, output_file)
+
+        print("Audio generated successfully:", output_file)
+    else:
+        print("Failed to transcribe audio.")

speechtotext.py

@@ -0,0 +1,61 @@
+import streamlit as st
+from deepgram import Deepgram
+from transcript import deepgram
+import requests
+import urllib3
+import base64
+from transcript import summarize_text
+urllib3.disable_warnings()
+
+def speech_to_text():
+    st.title("Speech to Text")
+
+    # File uploader for audio files
+    uploaded_file = st.sidebar.file_uploader("Upload Audio File", type=["wav", "mp3"])
+
+    if uploaded_file is not None:
+        st.sidebar.audio(uploaded_file, format='audio/wav/mp3')  # Display the uploaded audio file
+        st.spinner("Uploading in progress....")
+        
+       # Language selection dropdown
+        options = st.sidebar.selectbox("Select Language", ["English","Hindi"])
+        if options=='English':
+            language="en"
+        elif options=='Hindi':
+            language="hi"
+    
+       
+
+        # Button to trigger transcription
+        if st.sidebar.button("Transcribe"):
+            with st.spinner("Please Wait,Transcribing in progress..."):
+                # Call the transcribe_audio function
+                transcript = deepgram(uploaded_file, language)
+                with st.container(height=500):
+                 st.markdown(transcript)
+                
+                st.download_button(
+                    label="Download Transcribed Text",
+                    data=transcript,
+                    file_name="transcribed_text.txt",
+                    mime="text/plain"
+                )                
+                
+            
+                        # Summarize transcription
+                summary = summarize_text(transcript)
+
+                # Display summary
+                st.subheader("Summary:")
+                st.write(summary)
+                st.download_button(
+                    label="Download Summary",
+                    data=summary,
+                    file_name="summary.txt",
+                    mime="text/plain",
+
+                )    
+if __name__ == "__main__":
+    speech_to_text()
+
+

t2s.py

@@ -0,0 +1,41 @@
+import streamlit as st
+from gtts import gTTS
+import os
+
+# Function to save the speech to a file using gTTS
+def save_speech(text, lang, filename):
+    tts = gTTS(text=text, lang=lang)
+    tts.save(filename)
+
+# Function to run the Streamlit app
+def text_to_speech_page():
+    st.title("Text-to-Speech")
+
+    # Dropdown to select language
+    language = st.selectbox("Select Language", ["English", "Hindi", "Punjabi"])
+
+    # Text input for the user to enter text
+    text_input = st.text_area("Enter text to convert to speech", "")
+
+    # Button to trigger TTS
+    if st.button("Convert to Speech"):
+        if text_input:
+            # Map the selected language to gTTS language codes
+            lang_code = 'en'
+            if language == "Hindi":
+                lang_code = 'hi'
+            elif language == "Punjabi":
+                lang_code = 'pa'
+
+            # Save the speech to an audio file
+            output_audio_file = "output.mp3"
+            save_speech(text_input, lang_code, output_audio_file)
+
+            # Play the audio file
+            st.audio(output_audio_file, format='audio/mp3')
+        else:
+            st.warning("Please enter some text to convert.")
+
+# Run the app
+if __name__ == "__main__":
+    text_to_speech_page()

transcript.py

@@ -0,0 +1,47 @@
+
+from deepgram import (
+    DeepgramClient,
+    PrerecordedOptions,
+    FileSource
+)
+import httpx
+from sumy.parsers.plaintext import PlaintextParser
+from sumy.nlp.tokenizers import Tokenizer
+from sumy.summarizers.lsa import LsaSummarizer
+
+def deepgram(buffer_data,language):
+    # STEP 1 Create a Deepgram client using the API key
+    deepgram = DeepgramClient('5da1cbb9a0b27337e1ae7dc141b48c6775912ed2')
+
+    payload: FileSource = {
+        "buffer": buffer_data,
+    }
+
+    #STEP 2: Configure Deepgram options for audio analysis
+    options = PrerecordedOptions(
+        model="nova-2",
+        smart_format=True,
+        language=language,
+        diarize=True
+
+    )
+    myTimeout = httpx.Timeout(300, connect=10.0)
+    # STEP 3: Call the transcribe_file method with the text payload and options
+    response = deepgram.listen.prerecorded.v("1").transcribe_file(payload, options,timeout=myTimeout)
+    text=response['results']['channels'][0]['alternatives'][0]['paragraphs']['transcript']
+    return text
+
+def summarize_text(text):
+    # Initialize parser
+    parser = PlaintextParser.from_string(text, Tokenizer("english"))
+    
+    # Initialize LSA Summarizer
+    summarizer = LsaSummarizer()
+    
+    # Summarize the document with 3 sentences
+    summary = summarizer(parser.document, 3)
+    
+    # Convert summary sentences to string
+    summary_text = " ".join([str(sentence) for sentence in summary])
+    
+    return summary_text

transcription.json

@@ -0,0 +1,14 @@
+
+Speaker 0: Learn English intensely learn English faster by
+
+Speaker 1: practicing intense immersion. While it's fine to learn slowly sometimes, you can make tremendous progress by studying intensely for 1 month. Intensity focuses your brain and immerses it in English you'll learn more you'll remember more you'll speak better and you'll feel more confident if you study English intensely for 1 full month. Do this 3 to 5 times a year. How should you study intensely?
+
+I recommend 4 or more hours a day of English study. You should divide this time over the entire day so for example listen 1 hour in the morning 1 hour during lunch, 1 hour going home from work, and 1 hour before sleep. Here's the plan I recommend. Number 1, listen to the lessons for 2 hours each day during your intensity month. Focus especially on the main audio articles and on the mini story and POV lessons.
+
+These are the most powerful lessons and they are the most important. If you spend 2 hours every day focused on these lessons, your speaking and understanding will improve very quickly. Number 2, listen for 1 hour a day to a movie. Use my movie technique. Study only one scene for 3 to 7 days.
+
+1st, use the subtitles to understand the vocabulary and meaning. Then, listen and use subtitles at the same time finally, turn off the subtitles and listen to the scene many times. And number 3, read an easy novel, 1 hour every day. I recommend children's novels to start with. Then you can gradually find more difficult books.
+
+You should not need a dictionary to understand the book. If you need a dictionary, it's too difficult, so find an easier book. By following the above plan for 30 days or more, you will create a sudden, big, powerful improvement in your English ability. Your listening and speaking will show a huge improvement. Your vocabulary will grow quickly you'll feel more confident when you speak you don't always need to follow this plan but try to use it for just 30 days.
+
+Do these intense study months every 2 or 3 months and you will get amazing results.