Update app.py

app.py

@@ -2,83 +2,94 @@ import streamlit as st
 from transformers import pipeline
 from PIL import Image
 
-# Cache model loading for performance
-@st.cache_resource
-def load_image_to_text():
-    return pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
-
-@st.cache_resource
-def load_story_generator():
-    return pipeline("text-generation", model="pranavpsv/genre-story-generator-v2")
-
-@st.cache_resource
-def load_tts():
-    return pipeline("text-to-speech", model="facebook/mms-tts-eng")
-
-# Generation functions with error handling
+# Function to generate a caption from an uploaded image
+# Uses the Hugging Face pipeline with the Salesforce BLIP model to analyze the image and produce a descriptive caption
 def generate_caption(image):
-    image_to_text = load_image_to_text()
-    try:
-        caption = image_to_text(image)[0]["generated_text"]
-        return caption
-    except Exception as e:
-        st.error(f"Oops! Something went wrong while generating the caption: {e}")
-        return None
+    image_to_text = pipeline("image-to-text", model="Salesforce/blip-image-captioning-base")
+    caption = image_to_text(image)[0]["generated_text"]
+    return caption
 
+# Function to generate a story based on the image caption
+# Uses the pranavpsv/genre-story-generator-v2 model to create a creative story inspired by the caption
 def generate_story(caption):
-    story_generator = load_story_generator()
-    try:
-        prompt = f"Once upon a time, there was {caption}. "
-        story = story_generator(prompt, max_length=200, do_sample=True, temperature=0.7)[0]['generated_text']
-        return story
-    except Exception as e:
-        st.error(f"Oops! Something went wrong while generating the story: {e}")
-        return None
+    pipe = pipeline("text-generation", model="pranavpsv/genre-story-generator-v2")
+    story = pipe(caption, max_length=150, do_sample=True)[0]['generated_text']  # Added max_length for shorter, kid-friendly stories
+    return story
 
+# Function to convert the story into audio
+# Uses the facebook/mms-tts-eng model to generate a spoken version of the story
 def generate_audio(story):
-    tts = load_tts()
-    try:
-        audio = tts(story)
-        return audio
-    except Exception as e:
-        st.error(f"Oops! Something went wrong while generating the audio: {e}")
-        return None
+    pipe = pipeline("text-to-speech", model="facebook/mms-tts-eng")
+    audio = pipe(story)
+    return audio
 
-# Streamlit UI
-st.markdown("<h1 style='text-align: center; color: blue;'>📸✨ Storyteller for Kids! ✨📸</h1>", unsafe_allow_html=True)
-st.markdown("<p style='text-align: center;'>Upload a fun picture and I’ll tell you a magical story about it!</p>", unsafe_allow_html=True)
+# Streamlit UI: Creating a fun and interactive interface for kids
+# The app is designed to be simple, colorful, and engaging for young users
 
-# File uploader for image
-uploaded_file = st.file_uploader("Choose an image", type=["png", "jpg", "jpeg"])
+# Display a vibrant, centered title with emojis to grab attention
+st.markdown("<h1 style='text-align: center; color: purple;'>🌟 Picture to Story Magic! 🌟</h1>", unsafe_allow_html=True)
 
+# Add a welcoming description to explain the app’s purpose in kid-friendly language
+st.markdown(
+    """
+    <p style='text-align: center; font-size: 16px;'>
+    🎉 Hello, young adventurers! Upload a picture, and I’ll create a magical story just for you! 
+    This app is perfect for kids aged 3-10 who love fun and creative storytelling. 
+    Let’s make some story magic together! 🧙‍♂️
+    </p>
+    """, 
+    unsafe_allow_html=True
+)
+
+# File uploader for images, with a playful prompt
+uploaded_file = st.file_uploader("📸 Pick a cool picture!", type=["png", "jpg", "jpeg"], help="Choose a fun image to start your story adventure!")
+
+# Check if an image has been uploaded
 if uploaded_file is not None:
-    # Display the uploaded image
+    # Load and display the uploaded image
+    # The image is opened using Pillow and displayed with a caption
     image = Image.open(uploaded_file)
-    st.image(image, caption="Your Uploaded Image", use_column_width=True)
-
-    # Generate and display caption with spinner
-    with st.spinner("Creating a caption..."):
-        generated_caption = generate_caption(image)
+    st.image(image, caption="Your Awesome Picture! 😊", use_column_width=True)  # Changed to use_column_width for better scaling
 
-    if generated_caption:
-        # Allow user to edit the caption
-        caption_input = st.text_area("Caption:", value=generated_caption, height=100)
-        st.write("Feel free to change the caption to make your own story!")
+    # Generate and display the image caption
+    # The caption describes what’s in the image, serving as the story’s starting point
+    with st.spinner("🔍 Looking at your picture..."):
+        image_caption = generate_caption(image)
+    
+    st.subheader("✨ What’s in Your Picture?")
+    st.write(f"{image_caption}")
+    st.markdown("<p style='font-size: 14px;'>This is what I see in your picture! It’s the start of our story! 📖</p>", unsafe_allow_html=True)
 
-        # Generate story when button is clicked
-        if st.button("Generate Story"):
-            with st.spinner("Writing a magical story..."):
-                story = generate_story(caption_input)
+    # Generate and display the story
+    # The story is created based on the caption, designed to be short and engaging for kids
+    with st.spinner("🖌️ Writing a magical story..."):
+        story_telling = generate_story(image_caption)
+    
+    st.subheader("📚 Your Magical Story!")
+    st.write(f"{story_telling}")
+    st.markdown("<p style='font-size: 14px;'>Wow, what a fun story! Want to hear it out loud? Click below! 🎶</p>", unsafe_allow_html=True)
 
-            if story:
-                st.subheader("Your Story:")
-                st.write(story)
-                st.download_button("Download Story", story, file_name="my_story.txt")
+    # Generate and play audio for the story
+    # Audio is generated only when the user clicks the button, keeping the app responsive
+    if st.button("🎙️ Listen to Your Story!"):
+        with st.spinner("🎵 Turning your story into sound..."):
+            audio = generate_audio(story_telling)
+        st.audio(
+            audio['audio'], 
+            format="audio/wav", 
+            start_time=0, 
+            sample_rate=audio['sampling_rate']
+        )
+        st.markdown("<p style='font-size: 14px;'>Amazing! You can listen to your story anytime! 😄</p>", unsafe_allow_html=True)
 
-                # Generate audio on demand
-                if st.button("Generate Audio"):
-                    with st.spinner("Turning your story into sound..."):
-                        audio = generate_audio(story)
-                    if audio:
-                        st.audio(audio['audio'], format="audio/wav", start_time=0, sample_rate=audio['sampling_rate'])
-                        st.download_button("Download Audio", audio['audio'], file_name="my_story.wav", mime="audio/wav")
+# Add a footer to encourage users to try again
+st.markdown(
+    """
+    <hr>
+    <p style='text-align: center; font-size: 14px;'>
+    🖼️ Want another adventure? Upload a new picture and create a brand-new story! 
+    Made with love for curious kids! ❤️
+    </p>
+    """, 
+    unsafe_allow_html=True
+)