Update app.py

app.py

@@ -1,7 +1,63 @@
-import streamlit as st 
-import keras
-import matplotlib.pyplot as plt 
-from keras.datasets import mnist
-from sklearn.preprocessing import MinMaxScaler
-from keras.models import Sequential,Model
-from keras.layers import Dense,ConV2D,MaxPooling2D,AveragePooling2D,InputLayer,Flatten
+import streamlit as st
+import matplotlib.pyplot as plt
+import numpy as np
+from tensorflow.keras.models import Model, load_model
+from tensorflow.keras.layers import Conv2D
+from PIL import Image
+import cv2
+
+st.title("CNN Layer Visualization (Upload Model & Image)")
+
+col1, col2 = st.columns([1, 2])
+
+with col1:
+    st.header("Upload Files")
+
+    # Upload model file
+    model_file = st.file_uploader("Upload a Keras model (.h5 or .keras)", type=["h5", "keras"])
+
+    # Upload image file
+    uploaded_file = st.file_uploader("Upload an image", type=["jpg", "jpeg", "png"])
+    
+    if model_file is not None and uploaded_file is not None:
+        # Load model
+        model = load_model(model_file)
+
+        # Show uploaded image
+        image = Image.open(uploaded_file).convert("RGB")
+        st.image(image, caption="Uploaded Image", use_column_width=True)
+        
+        # Preprocess image
+        img_array = np.array(image)
+        img_resized = cv2.resize(img_array, (128, 128)) / 255.0
+        img_input = np.expand_dims(img_resized, axis=0)
+
+        # Collect conv layer outputs
+        conv_layers = [layer for layer in model.layers if isinstance(layer, Conv2D)]
+        outputs = [layer.output for layer in conv_layers]
+        feature_model = Model(inputs=model.input, outputs=outputs)
+        feature_maps = feature_model.predict(img_input)
+
+with col2:
+    st.header("Layer Output Visualizer")
+    if model_file is not None and uploaded_file is not None:
+        conv_layer_names = [layer.name for layer in model.layers if isinstance(layer, Conv2D)]
+
+        # Loop through each conv layer
+        for fmap, lname in zip(feature_maps, conv_layer_names):
+            num_filters = fmap.shape[-1]
+            cols = 8
+            rows = num_filters // cols if num_filters % cols == 0 else num_filters // cols + 1
+
+            fig, axes = plt.subplots(rows, cols, figsize=(15, 15))
+            fig.suptitle(f"Layer: {lname}", fontsize=16)
+
+            for i in range(rows * cols):
+                if i < num_filters:
+                    ax = axes[i // cols, i % cols]
+                    ax.imshow(fmap[0, :, :, i], cmap="viridis")
+                    ax.axis("off")
+                else:
+                    axes[i // cols, i % cols].axis("off")
+
+            st.pyplot(fig)