app.py

import streamlit as st
from transformers import pipeline
from PIL import Image
import io
from gtts import gTTS
import time

# Set page title
st.set_page_config(page_title="Kids Story Generator")

# Title and introduction
st.title("Kids Story Generator")
st.write("Upload a picture and let's create a magical story!")

# Initialize models
@st.cache_resource
def load_models():
    image_to_text = pipeline("image-to-text", model="microsoft/git-base-coco")
    story_generator = pipeline("text-generation", model="gpt2")
    return image_to_text, story_generator

image_to_text, story_generator = load_models()

# Function to generate caption from image
def generate_caption(image):
    try:
        caption = image_to_text(image)[0]['generated_text']
        return caption
    except Exception as e:
        st.error(f"Error generating caption: {str(e)}")
        return "children playing in a colorful park"

# Function to generate story from caption (less than 100 words)
def generate_story(caption):
    try:
        prompt = f"Once upon a time, {caption} "
        
        # Set parameters for faster generation while keeping quality
        story = story_generator(
            prompt, 
            max_length=100, 
            do_sample=True,
            temperature=0.7,  # Lower temperature for faster generation
            top_p=0.9,
            num_return_sequences=1
        )[0]['generated_text']
        
        # Ensure story doesn't exceed 100 words
        words = story.split()
        if len(words) > 100:
            words = words[:100]
            story = " ".join(words)
            # Add period to the end if needed
            if not story.endswith(('.', '!', '?')):
                story += '.'
        
        return story
    except Exception as e:
        st.error(f"Error generating story: {str(e)}")
        return f"Once upon a time, {caption}. Something magical happened and everyone lived happily ever after."

# Function to convert text to speech
def text_to_speech(text):
    try:
        tts = gTTS(text=text, lang='en', slow=False)
        audio_file = "story_audio.mp3"
        tts.save(audio_file)
        return audio_file
    except Exception as e:
        st.error(f"Error generating audio: {str(e)}")
        return None

# File uploader
uploaded_file = st.file_uploader("Choose an image...", type=["jpg", "jpeg", "png"])

if uploaded_file is not None:
    try:
        # Display the uploaded image
        image = Image.open(uploaded_file)
        st.image(image, caption='Uploaded Image', use_container_width=True)
        
        # Generate button
        if st.button("Generate Story"):
            # Use progress bar for better UX
            progress_bar = st.progress(0)
            
            # Generate caption
            progress_bar.progress(25)
            st.text('Analyzing image...')
            caption = generate_caption(image)
            st.write("Image caption:", caption)
            
            # Generate story
            progress_bar.progress(50)
            st.text('Creating story...')
            story = generate_story(caption)
            word_count = len(story.split())
            st.write(f"### Your Story ({word_count} words)")
            st.write(story)
            
            # Generate audio
            progress_bar.progress(75)
            st.text('Generating audio...')
            audio_file = text_to_speech(story)
            
            # Display audio
            if audio_file:
                progress_bar.progress(100)
                st.write("### Listen to your story")
                st.audio(audio_file)
            
            # Clear progress when done
            progress_bar.empty()
    
    except Exception as e:
        st.error(f"An error occurred: {str(e)}")

st.markdown("---")
st.write("Created for ISOM5240 Assignment")