Update app.py

app.py

@@ -29,7 +29,7 @@ from prediction import (
 def overview_page():
     col1, col2, col3 = st.columns([1, 6, 1])
     
-    st.title("🚀 Beyond Normals:  Unveiling Image Anomalies with AI")
+    st.title("🚀 Beyond Normal:  Unveiling Image Anomalies with AI")
     st.markdown("---")
 
     try:
@@ -89,12 +89,7 @@ def overview_page():
 
     st.subheader("✨ **Bringing Anomalies to Light**: Real-World Examples")
     st.write("""
-        Below are **three real examples** of AI-powered anomaly detection. Each image set follows this pattern:
-        
-        🔹 **First Column**: Original object with an anomaly  
-        🔹 **Second Column**: AI-generated **heatmap** highlighting the defect  
-        🔹 **Third Column**: Segmentation mask identifying **defect locations**  
-        🔹 **Fourth Column**: **Ground truth** (expert-labeled defect areas for verification)  
+        Below are **three real examples** of AI-powered anomaly detection.         
     """)
 
     # Display anomaly detection images directly without additional data processing
@@ -103,42 +98,21 @@ def overview_page():
     st.image("images/anomaly_visual_example_3.png", use_container_width=True, caption="Leather - Cut Defect")
 
     st.write("""
-        🔵 **The Heatmap (Second Column)**:  
+         **First Column**: 
+        - Original object with an anomaly  
+             
+        **The Heatmap (Second Column)**:  
         - AI scans the object and **highlights unusual areas** in **red/yellow**, indicating anomaly.  
 
-        ⚫ **Segmentation Map (Third Column)**:  
+        **Segmentation Map (Third Column)**:  
         - Shows the **exact shape** of the detected anomaly, crucial for precise localization.
 
-        ✅ **Ground Truth (Fourth Column)**:  
+        **Ground Truth (Fourth Column)**:  
         - The manually labeled anomaly **used for AI validation**.
 
         This technology helps manufacturers **automate anomaly detection, reduce waste, and ensure top-tier product quality** at an industrial scale. 🚀  
     """)
 
-    st.success("🌟 With AI-powered anomaly detection, we bring **precision and automation** to manufacturing quality control! 🔍✨")
-
-    st.markdown("---")
-    col1, col2 = st.columns(2)
-    
-    with col1:
-        st.subheader("🎯 Process Pipeline")
-        st.markdown("""
-            1. **Data Collection & Preprocessing**
-            2. **Feature Extraction**
-            3. **Model Training**
-            4. **Anomaly Detection**
-            5. **Results Visualization**
-        """)
-    
-    with col2:
-        st.subheader("🛠️ Key Technologies")
-        st.markdown("""
-            - **Deep Learning**: PyTorch
-            - **Computer Vision**: OpenCV, PIL
-            - **Data Analysis**: NumPy, Pandas
-            - **Visualization**: Plotly, Matplotlib
-        """)
-
 
 # Dataset Page
 def dataset_page():
@@ -406,6 +380,8 @@ def resnet50_page():
         Essentially, these **1536 features** act like a summary of the image’s most important elements. 📝
 
         """)
+    st.image("images/internal_features.png", caption="ResNet50 Block 1 & 2, random internal features", use_container_width=True)
+    
 def models_page():
     st.title("🤖 Anomaly Detection Models")