Add ConvNeXt Large model and app files

allergen_data.json

@@ -0,0 +1,103 @@
+{
+  "Apple Pie": [],
+  "Baby Back Ribs": [],
+  "Baklava": ["Gluten", "Nuts"],
+  "Beef Carpaccio": [],
+  "Beef Tartare": ["Egg"],
+  "Beet Salad": [],
+  "Beignets": ["Gluten", "Dairy"],
+  "Bibimbap": ["Soy", "Egg"],
+  "Bread Pudding": ["Gluten", "Dairy", "Egg"],
+  "Breakfast Burrito": ["Gluten", "Dairy", "Egg"],
+  "Bruschetta": ["Gluten"],
+  "Caesar Salad": ["Dairy", "Fish"],
+  "Cannoli": ["Gluten", "Dairy"],
+  "Caprese Salad": ["Dairy"],
+  "Carrot Cake": ["Gluten", "Dairy", "Egg", "Nuts"],
+  "Ceviche": ["Fish", "Shellfish"],
+  "Cheesecake": ["Dairy", "Gluten", "Egg"],
+  "Cheese Plate": ["Dairy"],
+  "Chicken Curry": [],
+  "Chicken Quesadilla": ["Gluten", "Dairy"],
+  "Chicken Wings": [],
+  "Chocolate Cake": ["Gluten", "Dairy", "Egg"],
+  "Chocolate Mousse": ["Dairy", "Egg"],
+  "Churros": ["Gluten", "Dairy"],
+  "Clam Chowder": ["Shellfish", "Dairy", "Gluten"],
+  "Club Sandwich": ["Gluten", "Dairy", "Egg"],
+  "Crab Cakes": ["Shellfish", "Gluten", "Egg"],
+  "Creme Brulee": ["Dairy", "Egg"],
+  "Croque Madame": ["Gluten", "Dairy", "Egg"],
+  "Cup Cakes": ["Gluten", "Dairy", "Egg"],
+  "Deviled Eggs": ["Egg"],
+  "Donuts": ["Gluten", "Dairy", "Egg"],
+  "Dumplings": ["Gluten", "Soy"],
+  "Edamame": ["Soy"],
+  "Eggs Benedict": ["Gluten", "Dairy", "Egg"],
+  "Escargots": ["Shellfish", "Dairy"],
+  "Falafel": ["Gluten"],
+  "Filet Mignon": [],
+  "Fish And Chips": ["Fish", "Gluten"],
+  "Foie Gras": [],
+  "French Fries": [],
+  "French Onion Soup": ["Gluten", "Dairy"],
+  "French Toast": ["Gluten", "Dairy", "Egg"],
+  "Fried Calamari": ["Shellfish", "Gluten"],
+  "Fried Rice": ["Soy", "Egg"],
+  "Frozen Yogurt": ["Dairy"],
+  "Garlic Bread": ["Gluten", "Dairy"],
+  "Gnocchi": ["Gluten", "Dairy", "Egg"],
+  "Greek Salad": ["Dairy"],
+  "Grilled Cheese Sandwich": ["Gluten", "Dairy"],
+  "Grilled Salmon": ["Fish"],
+  "Guacamole": [],
+  "Gyoza": ["Gluten", "Soy"],
+  "Hamburger": ["Gluten"],
+  "Hot And Sour Soup": ["Soy", "Egg"],
+  "Hot Dog": ["Gluten"],
+  "Huevos Rancheros": ["Egg"],
+  "Hummus": ["Sesame"],
+  "Ice Cream": ["Dairy"],
+  "Lasagna": ["Gluten", "Dairy", "Egg"],
+  "Lobster Bisque": ["Shellfish", "Dairy"],
+  "Lobster Roll Sandwich": ["Shellfish", "Gluten", "Dairy"],
+  "Macaroni And Cheese": ["Gluten", "Dairy"],
+  "Macarons": ["Nuts", "Egg", "Dairy"],
+  "Miso Soup": ["Soy", "Fish"],
+  "Mussels": ["Shellfish"],
+  "Nachos": ["Dairy"],
+  "Omelette": ["Egg", "Dairy"],
+  "Onion Rings": ["Gluten"],
+  "Oysters": ["Shellfish"],
+  "Pad Thai": ["Peanuts", "Soy", "Egg", "Shellfish"],
+  "Paella": ["Shellfish", "Fish"],
+  "Pancakes": ["Gluten", "Dairy", "Egg"],
+  "Panna Cotta": ["Dairy"],
+  "Peking Duck": ["Gluten", "Soy"],
+  "Pho": [],
+  "Pizza": ["Gluten", "Dairy"],
+  "Pork Chop": [],
+  "Poutine": ["Dairy"],
+  "Prime Rib": [],
+  "Pulled Pork Sandwich": ["Gluten"],
+  "Ramen": ["Gluten", "Soy", "Egg"],
+  "Ravioli": ["Gluten", "Dairy", "Egg"],
+  "Red Velvet Cake": ["Gluten", "Dairy", "Egg"],
+  "Risotto": ["Dairy"],
+  "Samosa": ["Gluten"],
+  "Sashimi": ["Fish"],
+  "Scallops": ["Shellfish"],
+  "Seaweed Salad": ["Sesame"],
+  "Shrimp And Grits": ["Shellfish", "Dairy"],
+  "Spaghetti Bolognese": ["Gluten", "Dairy"],
+  "Spaghetti Carbonara": ["Gluten", "Dairy", "Egg"],
+  "Spring Rolls": ["Gluten", "Soy"],
+  "Steak": [],
+  "Strawberry Shortcake": ["Gluten", "Dairy"],
+  "Sushi": ["Fish", "Soy"],
+  "Tacos": [],
+  "Takoyaki": ["Gluten", "Shellfish", "Egg"],
+  "Tiramisu": ["Gluten", "Dairy", "Egg"],
+  "Tuna Tartare": ["Fish", "Soy"],
+  "Waffles": ["Gluten", "Dairy", "Egg"]
+} 

app.py

@@ -0,0 +1,581 @@
+import streamlit as st
+from PIL import Image
+import torch
+from torchvision import models
+import torch.nn as nn
+import torchvision.transforms as transforms
+import io
+import os
+import openai
+import json
+import timm  # For ConvNeXt Large model
+
+# --- Page Configuration ---
+st.set_page_config(page_title="EatSmart Pro", page_icon="🍽️", layout="wide")
+
+# Mobile menu indicator
+st.markdown("""
+<div style="display: block; background: linear-gradient(90deg, #667eea 0%, #764ba2 100%); 
+           padding: 8px 15px; border-radius: 8px; margin-bottom: 15px; text-align: center;">
+    <p style="color: white; margin: 0; font-size: 14px;">
+        📱 <strong>Mobile Users:</strong> Click the <strong>></strong> arrow (top-left) to open preferences menu
+    </p>
+</div>
+""", unsafe_allow_html=True)
+
+# ==============================================================================
+# The correct, full list of 101 class names from your training script.
+# ==============================================================================
+CLASS_NAMES = [
+    'apple_pie', 'baby_back_ribs', 'baklava', 'beef_carpaccio', 'beef_tartare',
+    'beet_salad', 'beignets', 'bibimbap', 'bread_pudding', 'breakfast_burrito',
+    'bruschetta', 'caesar_salad', 'cannoli', 'caprese_salad', 'carrot_cake',
+    'ceviche', 'cheesecake', 'cheese_plate', 'chicken_curry', 'chicken_quesadilla',
+    'chicken_wings', 'chocolate_cake', 'chocolate_mousse', 'churros', 'clam_chowder',
+    'club_sandwich', 'crab_cakes', 'creme_brulee', 'croque_madame', 'cup_cakes',
+    'deviled_eggs', 'donuts', 'dumplings', 'edamame', 'eggs_benedict',
+    'escargots', 'falafel', 'filet_mignon', 'fish_and_chips', 'foie_gras',
+    'french_fries', 'french_onion_soup', 'french_toast', 'fried_calamari', 'fried_rice',
+    'frozen_yogurt', 'garlic_bread', 'gnocchi', 'greek_salad', 'grilled_cheese_sandwich',
+    'grilled_salmon', 'guacamole', 'gyoza', 'hamburger', 'hot_and_sour_soup',
+    'hot_dog', 'huevos_rancheros', 'hummus', 'ice_cream', 'lasagna',
+    'lobster_bisque', 'lobster_roll_sandwich', 'macaroni_and_cheese', 'macarons', 'miso_soup',
+    'mussels', 'nachos', 'omelette', 'onion_rings', 'oysters',
+    'pad_thai', 'paella', 'pancakes', 'panna_cotta', 'peking_duck',
+    'pho', 'pizza', 'pork_chop', 'poutine', 'prime_rib',
+    'pulled_pork_sandwich', 'ramen', 'ravioli', 'red_velvet_cake', 'risotto',
+    'samosa', 'sashimi', 'scallops', 'seaweed_salad', 'shrimp_and_grits',
+    'spaghetti_bolognese', 'spaghetti_carbonara', 'spring_rolls', 'steak', 'strawberry_shortcake',
+    'sushi', 'tacos', 'takoyaki', 'tiramisu', 'tuna_tartare',
+    'waffles'
+]
+
+# ==============================================================================
+# ALL HELPER FUNCTIONS ARE NOW INSIDE A SINGLE APP.PY
+# ==============================================================================
+
+def get_convnext_model(num_classes):
+    """
+    Creates the ConvNeXt Large model architecture,
+    matching the training script for maximum accuracy.
+    """
+    print(f"🚀 Loading ConvNeXt Large model for inference...")
+    print(f"📊 Model: ConvNeXt Large (197M parameters)")
+    
+    # Create ConvNeXt Large model (same as training script)
+    model = timm.create_model('convnext_large.fb_in22k_ft_in1k', pretrained=True, num_classes=num_classes)
+    
+    # Model statistics for user feedback
+    total_params = sum(p.numel() for p in model.parameters())
+    print(f"✅ Model architecture loaded:")
+    print(f"   Total parameters: {total_params:,}")
+    print(f"   Model size: ~{total_params * 4 / 1024**2:.1f} MB")
+    
+    return model
+
+def get_efficientnet_model(num_classes):
+    """
+    Creates the old EfficientNet-V2-S model architecture for fallback.
+    """
+    print(f"⚠️ Loading EfficientNet-V2-S model (fallback)...")
+    model = models.efficientnet_v2_s(weights=None)
+    num_features = model.classifier[1].in_features
+    model.classifier = nn.Sequential(
+        nn.Dropout(p=0.2),
+        nn.Linear(num_features, 256),
+        nn.ReLU(),
+        nn.Dropout(p=0.1),
+        nn.Linear(256, num_classes)
+    )
+    return model
+
+def load_json_data(path):
+    if not os.path.exists(path): return {}
+    try:
+        with open(path, 'r') as f: return json.load(f)
+    except (FileNotFoundError, json.JSONDecodeError): return {}
+
+def get_health_info(food_name, health_data):
+    """Enhanced health information display with comprehensive nutritional data"""
+    food_name_key = food_name.replace('_', ' ').title()
+    
+    # Try to get from the nested nutrition_info structure first
+    nutrition_info = health_data.get("nutrition_info", {})
+    info = nutrition_info.get(food_name_key) or nutrition_info.get(food_name.lower())
+    
+    # Try direct lookup if nested lookup fails
+    if not info:
+        info = health_data.get(food_name_key)
+    
+    if not info:
+        return "<p>No specific health information available for this dish.</p>"
+    
+    # Get health score
+    health_scores = health_data.get("health_scores", {})
+    health_score = health_scores.get(food_name.lower(), {})
+    score = health_score.get("score", 75)
+    score_explanation = health_score.get("explanation", "Good")
+    
+    # Determine score color
+    if score >= 80:
+        score_color = "#28a745"  # Green
+        score_bg = "#d4edda"
+    elif score >= 60:
+        score_color = "#ffc107"  # Yellow
+        score_bg = "#fff3cd"
+    else:
+        score_color = "#dc3545"  # Red
+        score_bg = "#f8d7da"
+    
+    # Build nutritional information
+    nutrition_html = f"""
+    <div style="background-color: #f8f9fa; padding: 15px; border-radius: 8px; margin: 10px 0;">
+        <h4 style="color: #495057; margin-top: 0;">📊 Nutritional Information</h4>
+        <div style="display: grid; grid-template-columns: repeat(4, 1fr); gap: 15px; text-align: center;">
+            <div>
+                <div style="font-size: 14px; color: #6c757d;">Calories</div>
+                <div style="font-size: 24px; font-weight: bold; color: #495057;">{info.get("calories", "N/A")}</div>
+                <div style="font-size: 12px; color: #6c757d;">kcal</div>
+            </div>
+            <div>
+                <div style="font-size: 14px; color: #6c757d;">Protein</div>
+                <div style="font-size: 24px; font-weight: bold; color: #495057;">{info.get("protein", "N/A")}</div>
+                <div style="font-size: 12px; color: #6c757d;">g</div>
+            </div>
+            <div>
+                <div style="font-size: 14px; color: #6c757d;">Carbs</div>
+                <div style="font-size: 24px; font-weight: bold; color: #495057;">{info.get("carbs", "N/A")}</div>
+                <div style="font-size: 12px; color: #6c757d;">g</div>
+            </div>
+            <div>
+                <div style="font-size: 14px; color: #6c757d;">Fat</div>
+                <div style="font-size: 24px; font-weight: bold; color: #495057;">{info.get("fat", "N/A")}</div>
+                <div style="font-size: 12px; color: #6c757d;">g</div>
+            </div>
+        </div>
+    </div>
+    """
+    
+    # Health Score
+    score_html = f"""
+    <div style="background-color: {score_bg}; border: 1px solid {score_color}; border-radius: 8px; padding: 15px; margin: 10px 0;">
+        <h4 style="color: {score_color}; margin-top: 0;">🏥 Health Assessment</h4>
+        <div style="background-color: {score_color}; color: white; padding: 10px; border-radius: 5px; text-align: center; font-weight: bold;">
+            Health Score: {score}/100 ({score_explanation})
+        </div>
+    </div>
+    """
+    
+    # Health Benefits - Display using Streamlit components instead of HTML
+    benefits_section = ""
+    if info.get("benefits"):
+        # We'll handle benefits separately using st.markdown with proper formatting
+        benefits_section = "BENEFITS_SECTION"  # Placeholder to indicate benefits exist
+    
+    return nutrition_html + score_html + benefits_section
+
+def get_allergen_info(food_name, allergen_data):
+    """Enhanced allergen information using sidebar preferences"""
+    food_name_key = food_name.replace('_', ' ').title()
+    allergens = allergen_data.get(food_name_key, [])
+    
+    # Display allergen information for the detected food
+    if allergens:
+        # Check for user-specified allergen matches
+        user_allergen_matches = [a for a in allergens if a in st.session_state.user_allergens]
+        
+        if user_allergen_matches:
+            # High priority alert for user's allergens
+            st.error(f"🚨 **CRITICAL ALLERGEN ALERT**: This dish contains **{', '.join(user_allergen_matches)}** which you've marked as allergens to avoid!")
+        
+        # Display allergens in a more user-friendly way
+        st.markdown("### ⚠️ Allergens Detected in This Food")
+        
+        # Create columns for allergen badges
+        cols = st.columns(min(len(allergens), 4))
+        for i, allergen in enumerate(allergens):
+            with cols[i % len(cols)]:
+                if allergen in st.session_state.user_allergens:
+                    # User's marked allergens - show as error
+                    st.error(f"🚨 {allergen}")
+                else:
+                    # Other allergens - show as info
+                    st.info(f"ℹ️ {allergen}")
+        
+        # Add explanation
+        if user_allergen_matches:
+            st.markdown("---")
+            st.markdown("🔴 **Red alerts** are for allergens you've marked in your preferences")
+        else:
+            st.markdown("---")
+            st.markdown("💙 **Blue badges** show allergens present in this food")
+            
+    else:
+        st.success("✅ No common allergens typically found in this dish.")
+
+def get_trans_fat_analysis(food_name, health_data):
+    """Enhanced trans fat analysis with user preferences"""
+    food_name_lower = food_name.lower().replace('_', ' ')
+    
+    # Get trans fat ingredients from health data
+    trans_fat_ingredients = health_data.get("trans_fat_ingredients", [
+        "hydrogenated oil", "partially hydrogenated oil", "margarine", "shortening"
+    ])
+    
+    # Foods that commonly contain trans fats
+    high_trans_fat_foods = [
+        "donuts", "french_fries", "fried", "margarine", "shortening", 
+        "processed", "packaged", "fast food", "baked goods", "cookies", "crackers"
+    ]
+    
+    # Check if food likely contains trans fats
+    likely_trans_fat = any(ingredient in food_name_lower for ingredient in high_trans_fat_foods)
+    
+    if likely_trans_fat:
+        alert_level = "HIGH RISK" if st.session_state.avoid_trans_fat else "POTENTIAL RISK"
+        if st.session_state.avoid_trans_fat:
+            st.error(f"🚨 **TRANS FAT ALERT**: You've enabled trans fat warnings and this food may contain trans fats!")
+        
+        trans_fat_html = f"""
+        <div style="background: linear-gradient(135deg, #f8d7da 0%, #f5c6cb 100%); border: 2px solid #dc3545; border-radius: 12px; padding: 20px; margin: 15px 0; box-shadow: 0 4px 6px rgba(0,0,0,0.1);">
+            <h4 style="color: #721c24; margin-top: 0; text-align: center;">🧪 Trans Fat Analysis - {alert_level}</h4>
+            <div style="color: #721c24; text-align: center;">
+                <p><strong>⚠️ This food may contain trans fats from:</strong></p>
+                <div style="display: flex; flex-wrap: wrap; justify-content: center; gap: 10px; margin: 15px 0;">
+                    {"".join([f'<span style="background-color: #dc3545; color: white; padding: 5px 10px; border-radius: 15px; font-size: 0.9em;">{ingredient.title()}</span>' for ingredient in trans_fat_ingredients])}
+                </div>
+                <p style="font-weight: bold; margin-top: 15px;">💡 <em>Recommendation: Check ingredient labels and consider healthier alternatives</em></p>
+            </div>
+        </div>
+        """
+    else:
+        trans_fat_html = f"""
+        <div style="background: linear-gradient(135deg, #d4edda 0%, #c3e6cb 100%); border: 2px solid #28a745; border-radius: 12px; padding: 20px; margin: 15px 0; box-shadow: 0 4px 6px rgba(0,0,0,0.1);">
+            <h4 style="color: #155724; margin-top: 0; text-align: center;">🧪 Trans Fat Analysis - LOW RISK</h4>
+            <div style="color: #155724; text-align: center;">
+                <p><strong>✅ Great Choice!</strong></p>
+                <p>This food typically contains minimal or no artificial trans fats.</p>
+                <p style="font-style: italic; margin-top: 10px;">Keep enjoying healthy foods like this! 🌟</p>
+            </div>
+        </div>
+        """
+    
+    return trans_fat_html
+
+def generate_recipe(food_name):
+    """Enhanced recipe generation based on detected ingredients and user preferences"""
+    try:
+        api_key = st.secrets.get("OPENAI_API_KEY")
+        if not api_key: 
+            return "Error: OPENAI_API_KEY not found."
+        
+        openai.api_key = api_key
+        
+        # Build prompt based on user preferences
+        dietary_restrictions = ""
+        if st.session_state.dietary_preferences:
+            dietary_restrictions = f"Make it {', '.join(st.session_state.dietary_preferences).lower()}. "
+        
+        allergen_restrictions = ""
+        if st.session_state.user_allergens:
+            allergen_restrictions = f"Avoid using {', '.join(st.session_state.user_allergens).lower()}. "
+        
+        trans_fat_note = ""
+        if st.session_state.avoid_trans_fat:
+            trans_fat_note = "Use healthy oils and avoid trans fats. "
+        
+        prompt = f"""Create a recipe for {food_name}. {dietary_restrictions}{allergen_restrictions}{trans_fat_note}
+        
+        Format the response with these exact headings:
+        Ingredients:
+        Instructions:
+        Chef's Tips:
+        
+        Make it practical for home cooking and include nutritional benefits."""
+        
+        response = openai.chat.completions.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "system", "content": "You are a professional chef who creates healthy, personalized recipes based on dietary preferences and restrictions."},
+                {"role": "user", "content": prompt}
+            ],
+            temperature=0.7, 
+            max_tokens=600
+        )
+        
+        recipe_text = response.choices[0].message.content
+        return recipe_text.replace("**Ingredients:**", "Ingredients:").replace("**Instructions:**", "Instructions:").replace("**Chef's Tips:**", "Chef's Tips:")
+    except Exception as e: 
+        return f"The AI Chef is busy right now. Error: {str(e)}"
+
+@st.cache_resource
+def load_model_resources():
+    try:
+        num_classes = len(CLASS_NAMES)
+        
+        # Look for the new ConvNeXt Large model first, fallback to old models
+        convnext_model_path = 'models/food_classifier_convnext_large_cpu_full.pth'
+        efficientnet_model_path = 'models/food101_efficientnet_best.pth'
+        old_model_path = 'models/final_model.pth'
+        
+        if os.path.exists(convnext_model_path):
+            model_path = convnext_model_path
+            model = get_convnext_model(num_classes=num_classes)
+            print(f"🎯 Using NEW ConvNeXt Large model: {model_path}")
+        elif os.path.exists(efficientnet_model_path):
+            model_path = efficientnet_model_path
+            model = get_efficientnet_model(num_classes=num_classes)
+            print(f"⚠️ Using EfficientNet model: {model_path}")
+        elif os.path.exists(old_model_path):
+            model_path = old_model_path
+            model = get_efficientnet_model(num_classes=num_classes)
+            print(f"⚠️ Using old fallback model: {model_path}")
+        else:
+            st.error(f"FATAL: No model file found. Looking for:\n- {convnext_model_path}\n- {efficientnet_model_path}\n- {old_model_path}")
+            return None, None, None, None
+        
+        device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
+        checkpoint = torch.load(model_path, map_location=device)
+        model.load_state_dict(checkpoint['model_state_dict'])
+        model.to(device)
+        model.eval()
+        
+        # Store model info for UI display
+        model_info = {
+            'path': model_path,
+            'type': 'ConvNeXt Large' if 'convnext' in model_path else 'EfficientNet-V2-S',
+            'accuracy': checkpoint.get('accuracy', 'Unknown')
+        }
+        
+        health_data = load_json_data('health_data.json')
+        allergen_data = load_json_data('allergen_data.json')
+        return model, health_data, allergen_data, model_info
+    except Exception as e:
+        st.error(f"A critical error occurred while loading the model: {e}")
+        return None, None, None, None
+
+model, health_data, allergen_data, model_info = load_model_resources()
+
+def transform_image(image_bytes):
+    transform = transforms.Compose([
+        transforms.Resize((224, 224)), transforms.ToTensor(),
+        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])])
+    image = Image.open(io.BytesIO(image_bytes)).convert("RGB")
+    return transform(image).unsqueeze(0)
+
+def get_prediction(image_tensor):
+    if model is None: return "Error: Model not loaded", 0.0
+    with torch.no_grad():
+        outputs = model(image_tensor.to(torch.device('cuda' if torch.cuda.is_available() else 'cpu')))
+        probabilities = torch.nn.functional.softmax(outputs, dim=1)
+        confidence, predicted_idx = torch.max(probabilities, 1)
+        predicted_idx_item = predicted_idx.item()
+        if predicted_idx_item >= len(CLASS_NAMES):
+            return "Prediction Error: Index out of bounds.", 0.0
+        predicted_class = CLASS_NAMES[predicted_idx_item].replace('_', ' ').title()
+        return predicted_class, confidence.item()
+
+# --- UI Layout ---
+st.markdown("""
+<div style="text-align: center; padding: 20px 0;">
+    <h1 style="font-size: 3em; margin: 0;">
+        🍽️ <span style="color: #28a745;">Eat</span><span style="color: #dc3545;">Smart</span> 
+        <span style="color: #17a2b8;">Pro</span>
+    </h1>
+    <p style="font-size: 1.2em; color: #6c757d; margin: 10px 0;">
+        🌟 Your <span style="color: #28a745;">AI-Powered</span> 
+        <span style="color: #dc3545;">Nutrition</span> Assistant 🌟
+    </p>
+    <div style="display: flex; justify-content: center; gap: 10px; margin: 15px 0;">
+        <span style="background: linear-gradient(45deg, #28a745, #20c997); color: white; padding: 5px 15px; border-radius: 20px; font-size: 0.9em;">
+            🥗 Healthy Analysis
+        </span>
+        <span style="background: linear-gradient(45deg, #dc3545, #fd7e14); color: white; padding: 5px 15px; border-radius: 20px; font-size: 0.9em;">
+            ⚠️ Allergen Alerts
+        </span>
+        <span style="background: linear-gradient(45deg, #17a2b8, #6f42c1); color: white; padding: 5px 15px; border-radius: 20px; font-size: 0.9em;">
+            🍳 Smart Recipes
+        </span>
+    </div>
+</div>
+""", unsafe_allow_html=True)
+
+# Model Status Indicator
+if model_info:
+    model_type = model_info['type']
+    model_accuracy = model_info['accuracy']
+    
+    if 'ConvNeXt' in model_type:
+        status_color = "#28a745"  # Green for new model
+        status_bg = "#d4edda"
+        status_icon = "🚀"
+        status_text = f"HIGH ACCURACY MODEL ACTIVE"
+    else:
+        status_color = "#ffc107"  # Yellow for old model
+        status_bg = "#fff3cd"
+        status_icon = "⚠️"
+        status_text = f"FALLBACK MODEL (Training in Progress)"
+    
+    st.markdown(f"""
+    <div style="background-color: {status_bg}; border: 2px solid {status_color}; border-radius: 10px; padding: 15px; margin: 15px 0; text-align: center;">
+        <div style="color: {status_color}; font-size: 1.2em; font-weight: bold;">
+            {status_icon} {status_text}
+        </div>
+        <div style="color: #495057; font-size: 0.9em; margin-top: 5px;">
+            Model: {model_type} | Validation Accuracy: {model_accuracy}
+        </div>
+    </div>
+    """, unsafe_allow_html=True)
+
+# User Preferences Sidebar
+with st.sidebar:
+    st.markdown("""
+    <div style="text-align: center; padding: 15px; background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); border-radius: 10px; margin-bottom: 20px;">
+        <h3 style="color: white; margin: 0;">⚙️ Your Preferences</h3>
+        <p style="color: #f8f9fa; margin: 5px 0; font-size: 0.9em;">Set your dietary needs</p>
+    </div>
+    """, unsafe_allow_html=True)
+    
+    # Initialize session state for preferences
+    if 'user_allergens' not in st.session_state:
+        st.session_state.user_allergens = []
+    if 'avoid_trans_fat' not in st.session_state:
+        st.session_state.avoid_trans_fat = False
+    if 'dietary_preferences' not in st.session_state:
+        st.session_state.dietary_preferences = []
+    
+    # Allergen Preferences
+    st.markdown("### 🚨 Allergen Alerts")
+    st.markdown("*Select allergens you want to be warned about:*")
+    
+    common_allergens = ["Gluten", "Dairy", "Egg", "Fish", "Shellfish", "Nuts", "Peanuts", "Soy", "Sesame"]
+    
+    for allergen in common_allergens:
+        if st.checkbox(f"🛡️ {allergen}", key=f"allergen_{allergen}", 
+                      value=allergen in st.session_state.user_allergens):
+            if allergen not in st.session_state.user_allergens:
+                st.session_state.user_allergens.append(allergen)
+        else:
+            if allergen in st.session_state.user_allergens:
+                st.session_state.user_allergens.remove(allergen)
+    
+    # Trans Fat Preference
+    st.markdown("### 🧪 Trans Fat Settings")
+    st.session_state.avoid_trans_fat = st.checkbox(
+        "⚠️ Alert me about trans fats", 
+        value=st.session_state.avoid_trans_fat,
+        help="Get warnings about foods that may contain trans fats"
+    )
+    
+    # Dietary Preferences
+    st.markdown("### 🌱 Dietary Preferences")
+    dietary_options = ["Vegetarian", "Vegan", "Keto", "Low-Carb", "High-Protein", "Gluten-Free"]
+    
+    for diet in dietary_options:
+        if st.checkbox(f"🥬 {diet}", key=f"diet_{diet}", 
+                      value=diet in st.session_state.dietary_preferences):
+            if diet not in st.session_state.dietary_preferences:
+                st.session_state.dietary_preferences.append(diet)
+        else:
+            if diet in st.session_state.dietary_preferences:
+                st.session_state.dietary_preferences.remove(diet)
+    
+    # Display current preferences summary
+    if st.session_state.user_allergens or st.session_state.dietary_preferences or st.session_state.avoid_trans_fat:
+        st.markdown("---")
+        st.markdown("### 📋 Active Preferences")
+        if st.session_state.user_allergens:
+            st.markdown(f"🚨 **Allergen Alerts:** {', '.join(st.session_state.user_allergens)}")
+        if st.session_state.dietary_preferences:
+            st.markdown(f"🌱 **Diet:** {', '.join(st.session_state.dietary_preferences)}")
+        if st.session_state.avoid_trans_fat:
+            st.markdown("🧪 **Trans Fat Alerts:** Enabled")
+
+if 'image_buffer' not in st.session_state: st.session_state.image_buffer = None
+if 'prediction_result' not in st.session_state: st.session_state.prediction_result = None
+if 'last_image_buffer' not in st.session_state: st.session_state.last_image_buffer = None
+
+col1, col2 = st.columns([1, 1.2])
+with col1:
+    st.markdown("""
+    <div style="text-align: center; padding: 15px; background: linear-gradient(135deg, #74b9ff 0%, #0984e3 100%); border-radius: 10px; margin-bottom: 20px;">
+        <h3 style="color: white; margin: 0;">📸 Upload Food Image</h3>
+        <p style="color: #ddd; margin: 5px 0; font-size: 0.9em;">Drag & drop or browse to analyze</p>
+    </div>
+    """, unsafe_allow_html=True)
+    
+    uploaded_file = st.file_uploader("Choose your food image", type=["jpg", "jpeg", "png"], label_visibility="collapsed")
+    
+    # Set the image buffer based on the file uploader's state.
+    if uploaded_file is not None:
+        st.session_state.image_buffer = uploaded_file.getvalue()
+    else:
+        st.session_state.image_buffer = None
+
+    # This code block displays the image after it is uploaded.
+    if st.session_state.image_buffer is not None:
+        st.image(st.session_state.image_buffer, caption='🍽️ Your Food Image', use_column_width=True)
+
+with col2:
+    st.markdown("""
+    <div style="text-align: center; padding: 15px; background: linear-gradient(135deg, #00b894 0%, #00a085 100%); border-radius: 10px; margin-bottom: 20px;">
+        <h3 style="color: white; margin: 0;">🔬 Smart Analysis & Recipes</h3>
+        <p style="color: #ddd; margin: 5px 0; font-size: 0.9em;">AI-powered nutrition insights</p>
+    </div>
+    """, unsafe_allow_html=True)
+    
+    if model and st.session_state.image_buffer:
+        if st.session_state.image_buffer != st.session_state.last_image_buffer:
+            st.session_state.last_image_buffer = st.session_state.image_buffer
+            with st.spinner('Analyzing image...'):
+                image_tensor = transform_image(st.session_state.image_buffer)
+                st.session_state.prediction_result = get_prediction(image_tensor)
+                if 'recipe' in st.session_state: del st.session_state.recipe
+        if st.session_state.prediction_result:
+            food_name, confidence = st.session_state.prediction_result
+            st.metric(label="Predicted Food", value=food_name)
+            st.progress(confidence, text=f"Confidence: {confidence:.2%}")
+            tab1, tab2, tab3, tab4 = st.tabs(["Health Info", "Allergen Alert", "Trans Fat Analysis", "AI Recipes"])
+            with tab1: 
+                health_info_html = get_health_info(food_name, health_data)
+                if "BENEFITS_SECTION" in health_info_html:
+                    # Display HTML parts
+                    st.markdown(health_info_html.replace("BENEFITS_SECTION", ""), unsafe_allow_html=True)
+                    
+                    # Display benefits using Streamlit components for proper formatting
+                    nutrition_info = health_data.get("nutrition_info", {})
+                    food_info = nutrition_info.get(food_name.replace('_', ' ').title()) or nutrition_info.get(food_name.lower()) or health_data.get(food_name.replace('_', ' ').title())
+                    
+                    if food_info and food_info.get("benefits"):
+                        st.markdown("### ✨ Health Benefits")
+                        for benefit in food_info.get("benefits", []):
+                            st.markdown(f"• {benefit}")
+                else:
+                    st.markdown(health_info_html, unsafe_allow_html=True)
+            with tab2: 
+                get_allergen_info(food_name, allergen_data)
+            with tab3:
+                st.markdown(get_trans_fat_analysis(food_name, health_data), unsafe_allow_html=True)
+            with tab4:
+                st.subheader(f"AI-Generated Recipe for {food_name}")
+                if 'recipe' not in st.session_state or st.session_state.get('recipe_food') != food_name:
+                    with st.spinner("Chef AI is thinking of a recipe..."):
+                        st.session_state.recipe = generate_recipe(food_name)
+                        st.session_state.recipe_food = food_name
+                if 'recipe' in st.session_state:
+                    recipe_text = st.session_state.recipe
+                    sections = {"ingredients": [], "instructions": [], "tips": []}
+                    current_section_key = None
+                    for line in recipe_text.split('\n'):
+                        line_lower = line.strip().lower()
+                        if line_lower.startswith("ingredients"): current_section_key = "ingredients"
+                        elif line_lower.startswith("instructions"): current_section_key = "instructions"
+                        elif line_lower.startswith("tips") or line_lower.startswith("chef's tips"): current_section_key = "tips"
+                        elif line.strip() and current_section_key: sections[current_section_key].append(line.strip().lstrip('*- '))
+                    with st.expander("Ingredients", expanded=True): st.markdown("\n".join(f"- {item}" for item in sections["ingredients"]) or "No ingredients listed.")
+                    with st.expander("Instructions", expanded=True): st.markdown("\n".join(f"{i+1}. {item}" for i, item in enumerate(sections["instructions"])) or "No instructions provided.")
+                    with st.expander("Chef's Tips"): st.markdown("\n".join(f"- {item}" for item in sections["tips"]) or "No special tips provided.")
+    elif not model:
+        st.error("Application has failed to start. Please check the logs for errors.")
+    else: 
+        st.info("Upload an image to get started.")

food_classifier_convnext_large_cpu_full.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9f23f0343f88bdcb948d9c4773943c2d4fdd841d745e8f2e6ec55d89848f755e
+size 2357079991

health_data.json

@@ -0,0 +1,73 @@
+{
+  "nutrition_info": {
+    "hamburger": {"calories": 550, "protein": 25, "carbs": 40, "fat": 30},
+    "pizza": {"calories": 285, "protein": 12, "carbs": 36, "fat": 10},
+    "french_fries": {"calories": 312, "protein": 3, "carbs": 41, "fat": 15},
+    "ice_cream": {"calories": 207, "protein": 3.5, "carbs": 24, "fat": 11},
+    "donuts": {"calories": 452, "protein": 5, "carbs": 51, "fat": 25},
+    "chocolate_cake": {"calories": 350, "protein": 5, "carbs": 45, "fat": 20},
+    "sushi": {"calories": 150, "protein": 8, "carbs": 28, "fat": 1},
+    "grilled_salmon": {"calories": 206, "protein": 22, "carbs": 0, "fat": 12},
+    "caesar_salad": {"calories": 480, "protein": 9, "carbs": 12, "fat": 45},
+    "pho": {"calories": 400, "protein": 20, "carbs": 50, "fat": 10},
+    "tacos": {"calories": 226, "protein": 12, "carbs": 24, "fat": 9},
+    "falafel": {"calories": 333, "protein": 13, "carbs": 32, "fat": 18},
+    "ramen": {"calories": 436, "protein": 15, "carbs": 62, "fat": 14},
+    "pancakes": {"calories": 227, "protein": 6, "carbs": 28, "fat": 10},
+    "waffles": {"calories": 291, "protein": 8, "carbs": 33, "fat": 14},
+    "steak": {"calories": 679, "protein": 48, "carbs": 0, "fat": 53},
+    "spaghetti_bolognese": {"calories": 670, "protein": 32, "carbs": 70, "fat": 28},
+    "Greek Salad": {
+      "calories": 250,
+      "fat": 22,
+      "protein": 5,
+      "carbs": 8,
+      "benefits": [
+        "Rich in healthy fats from olive oil and feta.",
+        "Good source of vitamins from fresh vegetables."
+      ]
+    },
+    "Grilled Salmon": {
+      "calories": 280,
+      "fat": 18,
+      "protein": 30,
+      "carbs": 0,
+      "benefits": [
+        "Excellent source of high-quality protein.",
+        "Rich in Omega-3 fatty acids, which support heart and brain health.",
+        "Contains Vitamin D and B vitamins."
+      ]
+    },
+    "Guacamole": {
+      "calories": 150,
+      "fat": 14,
+      "protein": 2,
+      "carbs": 10
+    }
+  },
+  "trans_fat_ingredients": [
+    "hydrogenated oil",
+    "partially hydrogenated oil",
+    "margarine",
+    "shortening"
+  ],
+  "health_scores": {
+    "hamburger": {"score": 70, "explanation": "Good"},
+    "pizza": {"score": 65, "explanation": "Average"},
+    "french_fries": {"score": 40, "explanation": "Poor"},
+    "ice_cream": {"score": 30, "explanation": "Poor"},
+    "donuts": {"score": 20, "explanation": "Poor"},
+    "chocolate_cake": {"score": 35, "explanation": "Poor"},
+    "sushi": {"score": 85, "explanation": "Excellent"},
+    "grilled_salmon": {"score": 95, "explanation": "Excellent"},
+    "caesar_salad": {"score": 55, "explanation": "Average"},
+    "pho": {"score": 80, "explanation": "Excellent"},
+    "tacos": {"score": 75, "explanation": "Good"},
+    "falafel": {"score": 80, "explanation": "Excellent"},
+    "ramen": {"score": 60, "explanation": "Average"},
+    "pancakes": {"score": 50, "explanation": "Average"},
+    "waffles": {"score": 45, "explanation": "Poor"},
+    "steak": {"score": 70, "explanation": "Good"},
+    "spaghetti_bolognese": {"score": 65, "explanation": "Average"}
+  }
+}