Spaces:
Runtime error
Runtime error
| import streamlit as st | |
| from PIL import Image | |
| import io | |
| import numpy as np | |
| # Import our custom modules | |
| from utils.image_preprocessing import preprocess_image | |
| from models.document_ai import extract_text_and_layout | |
| from models.text_processor import process_menu_text | |
| from models.braille_translator import text_to_braille, get_braille_metadata | |
| from utils.braille_display import create_braille_html, create_braille_comparison | |
| # App title and description | |
| st.title("Menu to Braille Converter") | |
| st.write("Upload a menu image to convert it to Braille text") | |
| # Sidebar for model settings | |
| st.sidebar.header("Settings") | |
| use_llm = st.sidebar.checkbox("Use LLM for text processing", value=True) | |
| use_context = st.sidebar.checkbox("Use AI for context enhancement", value=True) | |
| show_comparison = st.sidebar.checkbox("Show text/Braille comparison", value=True) | |
| # Add information about the application | |
| st.sidebar.markdown("---") | |
| st.sidebar.subheader("About") | |
| st.sidebar.info( | |
| "This application converts menu images to Braille text using AI. " | |
| "It extracts text from images using document AI, processes the text with LLMs, " | |
| "and converts to Braille." | |
| ) | |
| # File uploader | |
| uploaded_file = st.file_uploader("Choose a menu image...", type=["jpg", "jpeg", "png"]) | |
| # Display uploaded image and process it | |
| if uploaded_file is not None: | |
| # Load and display image | |
| image = Image.open(uploaded_file) | |
| st.image(image, caption="Uploaded Menu", use_column_width=True) | |
| # Add a button to process the image | |
| if st.button("Process Menu"): | |
| with st.spinner("Processing image..."): | |
| # Preprocess the image | |
| preprocessed_img = preprocess_image(image) | |
| # Extract text using LayoutLMv2 | |
| try: | |
| result = extract_text_and_layout(preprocessed_img) | |
| # Display extracted words | |
| if result['words']: | |
| raw_text = ' '.join(result['words']) | |
| # Show raw text in an expandable section | |
| with st.expander("Raw Extracted Text"): | |
| st.text_area("Raw OCR Output", raw_text, height=150) | |
| # Process text with LLM if enabled | |
| if use_llm: | |
| st.subheader("Processed Menu Text") | |
| with st.spinner("Enhancing text with AI..."): | |
| processed_result = process_menu_text(raw_text) | |
| if processed_result['success']: | |
| processed_text = processed_result['structured_text'] | |
| st.text_area("Structured Menu Text", processed_text, height=200) | |
| # Store the processed result for later use | |
| st.session_state.processed_text = processed_text | |
| st.session_state.menu_data = processed_result.get('menu_data', {}) | |
| else: | |
| st.warning(f"AI processing failed: {processed_result.get('error', 'Unknown error')}") | |
| processed_text = raw_text | |
| st.text_area("Text Output", processed_text, height=200) | |
| st.session_state.processed_text = processed_text | |
| else: | |
| # Just use the raw text | |
| st.subheader("Extracted Text") | |
| processed_text = raw_text | |
| st.text_area("Text Output", processed_text, height=200) | |
| st.session_state.processed_text = processed_text | |
| # Translate to Braille | |
| st.subheader("Braille Translation") | |
| with st.spinner("Translating to Braille..."): | |
| braille_result = text_to_braille(processed_text, use_context=use_context) | |
| if braille_result['success']: | |
| # Store for download | |
| st.session_state.braille_text = braille_result['formatted_braille'] | |
| # Display options | |
| display_option = st.radio( | |
| "Display format:", | |
| ["Text Only", "Visual Braille", "Side-by-Side Comparison"] | |
| ) | |
| if display_option == "Text Only": | |
| # Display Braille text as plain text | |
| st.text_area("Braille Output", braille_result['formatted_braille'], height=300) | |
| elif display_option == "Visual Braille": | |
| # Display Braille with visual representation | |
| braille_html = create_braille_html(braille_result['formatted_braille']) | |
| st.markdown(braille_html, unsafe_allow_html=True) | |
| else: # Side-by-Side Comparison | |
| # Display side-by-side comparison | |
| comparison_html = create_braille_comparison( | |
| processed_text, braille_result['formatted_braille'] | |
| ) | |
| st.markdown(comparison_html, unsafe_allow_html=True) | |
| # Display metadata | |
| metadata = get_braille_metadata(processed_text) | |
| st.info(f"Translation contains {metadata['word_count']} words, " | |
| f"{metadata['character_count']} characters, " | |
| f"{metadata['line_count']} lines.") | |
| # Show context summary if available | |
| if braille_result.get('context_summary'): | |
| with st.expander("AI Context Understanding"): | |
| st.write(braille_result['context_summary']) | |
| else: | |
| st.error(f"Braille translation failed: {braille_result.get('error', 'Unknown error')}") | |
| # Download options placeholder | |
| st.subheader("Download Options") | |
| st.info("PDF download will be implemented in Phase 5") | |
| else: | |
| st.warning("No text was extracted from the image.") | |
| except Exception as e: | |
| st.error(f"Error processing image: {str(e)}") | |