import os import re import smtplib import streamlit as st from dotenv import load_dotenv from transformers import pipeline from typing import Dict from gtts import gTTS from together import Together from email.mime.multipart import MIMEMultipart from email.mime.text import MIMEText from email.mime.audio import MIMEAudio # Load environment variables load_dotenv() # Image-to-text function def img2txt(url: str) -> str: print("Initializing captioning model...") captioning_model = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base") print("Generating text from the image...") text = captioning_model(url, max_new_tokens=20)[0]["generated_text"] print(text) return text # Text-to-story generation function def txt2story(prompt: str, top_k: int, top_p: float, temperature: float) -> str: client = Together(api_key=os.environ.get("TOGETHER_API_KEY")) story_prompt = f"Write a short story of no more than 250 words based on the following prompt: {prompt}" stream = client.chat.completions.create( model="meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", messages=[ {"role": "system", "content": '''As an experienced short story writer, write a meaningful story influenced by the provided prompt. Ensure the story does not exceed 250 words.'''}, {"role": "user", "content": story_prompt} ], top_k=top_k, top_p=top_p, temperature=temperature, stream=True ) story = '' for chunk in stream: story += chunk.choices[0].delta.content return story # Text-to-speech function def txt2speech(text: str) -> None: print("Converting text to speech using gTTS...") tts = gTTS(text=text, lang='en') tts.save("audio_story.mp3") # Get user preferences function def get_user_preferences() -> Dict[str, str]: preferences = { 'continent': st.selectbox("Continent", ["North America", "Europe", "Asia", "Africa", "Australia"]), 'genre': st.selectbox("Genre", ["Science Fiction", "Fantasy", "Mystery", "Romance"]), 'setting': st.selectbox("Setting", ["Future", "Medieval times", "Modern day", "Alternate reality"]), 'plot': st.selectbox("Plot", ["Hero's journey", "Solving a mystery", "Love story", "Survival"]), 'tone': st.selectbox("Tone", ["Serious", "Light-hearted", "Humorous", "Dark"]), 'theme': st.selectbox("Theme", ["Self-discovery", "Redemption", "Love", "Justice"]), 'conflict': st.selectbox("Conflict Type", ["Person vs. Society", "Internal struggle", "Person vs. Nature", "Person vs. Person"]), 'twist': st.selectbox("Mystery/Twist", ["Plot twist", "Hidden identity", "Unexpected ally/enemy", "Time paradox"]), 'ending': st.selectbox("Ending", ["Happy", "Bittersweet", "Open-ended", "Tragic"]) } return preferences # Function to send the story via email using smtplib def send_story_email(recipient_email: str, story_text: str, audio_file_path: str) -> bool: try: # Email configuration smtp_server = os.environ.get("SMTP_SERVER") # e.g., "smtp.gmail.com" smtp_port = int(os.environ.get("SMTP_PORT", 587)) # e.g., 587 for TLS sender_email = os.environ.get("SENDER_EMAIL") # Your email sender_password = os.environ.get("SENDER_PASSWORD") # Your email password # Create the email msg = MIMEMultipart() msg['From'] = sender_email msg['To'] = recipient_email msg['Subject'] = "Your Generated Story" # Attach the story text msg.attach(MIMEText(f"Here's your generated story:\n\n{story_text}\n\nEnjoy!", 'plain')) # Attach the audio file with open(audio_file_path, 'rb') as audio_file: audio_part = MIMEAudio(audio_file.read(), _subtype='mp3') audio_part.add_header('Content-Disposition', 'attachment', filename=os.path.basename(audio_file_path)) msg.attach(audio_part) # Send the email with smtplib.SMTP(smtp_server, smtp_port) as server: server.starttls() # Upgrade to a secure connection server.login(sender_email, sender_password) server.send_message(msg) return True except Exception as e: print(f"Error sending email: {str(e)}") return False # Basic email validation function def validate_email(email: str) -> bool: pattern = r'^[\w\.-]+@[\w\.-]+\.\w+$' return re.match(pattern, email) is not None # Main Streamlit application def main(): st.set_page_config( page_title="🎨 Image-to-Audio Story 🎧", page_icon="🖼️", layout="wide" ) st.title("Turn the Image into Audio Story") # Initialize session state variables if "story" not in st.session_state: st.session_state.story = "" if "audio_file_path" not in st.session_state: st.session_state.audio_file_path = "" if "caption" not in st.session_state: st.session_state.caption = "" # Main content area col1, col2 = st.columns([2, 3]) with col1: # Image upload section st.markdown("## 📷 Upload Image") uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"]) # Story preferences section st.markdown("## 🎭 Story Preferences") preferences = get_user_preferences() with col2: if uploaded_file is not None: # Display uploaded image st.markdown("## 🖼️ Your Image") bytes_data = uploaded_file.read() with open("uploaded_image.jpg", "wb") as file: file.write(bytes_data) st.image(uploaded_file, use_column_width=True) # Process image and generate story if st.button("🎨 Generate Story"): with st.spinner("🤖 AI is working its magic..."): try: # Get image description scenario = img2txt("uploaded_image.jpg") st.session_state.caption = scenario # Store caption in session state # Create story prompt prompt = f"""Based on the image description: '{scenario}', create a {preferences['genre']} story set in {preferences['setting']} in {preferences['continent']}. The story should have a {preferences['tone']} tone and explore the theme of {preferences['theme']}. The main conflict should be {preferences['conflict']}. The story should have a {preferences['twist']} and end with a {preferences['ending']} ending.""" # Generate story story = txt2story(prompt, top_k=5, top_p=0.8, temperature=1.5) st.session_state.story = story # Store story in session state # Convert to audio txt2speech(story) st.session_state.audio_file_path = "audio_story.mp3" # Store audio path in session state except Exception as e: st.error(f"An error occurred: {str(e)}") st.warning("Please try again or contact support if the problem persists.") # Display results if story exists in session state if st.session_state.story: st.markdown("---") # Image caption with st.expander("📜 Image Caption", expanded=True): st.write(st.session_state.caption) # Story text with st.expander("📖 Generated Story", expanded=True): st.write(st.session_state.story) # Audio player with st.expander("🎧 Audio Version", expanded=True): st.audio(st.session_state.audio_file_path) # Email section st.markdown("---") st.markdown("## 📧 Get Story via Email") email = st.text_input( "Enter your email address:", help="We'll send you the story text and audio file" ) if st.button("📤 Send to Email"): if not email: st.warning("Please enter an email address.") elif not validate_email(email): st.error("Please enter a valid email address.") else: print('The process is ongoing________________________') with st.spinner("📨 Sending email..."): if send_story_email(email, st.session_state.story, st.session_state.audio_file_path): st.success(f"Email sent to: {email}") else: st.error("❌ Failed to send email. Please try again.") if __name__ == '__main__': main()