Update app.py

app.py

@@ -1,4 +1,57 @@
 import streamlit as st
-# import train3
-x = st.slider('Select a value')
-st.write(x, 'squared is', x **x)
+import numpy as np
+from tensorflow.keras.preprocessing import image
+from tensorflow.keras.models import load_model
+
+# Define constants
+IMG_WIDTH, IMG_HEIGHT = 224, 224  # Adjust if your model requires different dimensions
+model_path = 'dog_cat_classifier_model.keras'  # Replace with your model path
+
+# Load the model (outside the main app function for efficiency)
+model = load_model(model_path)
+
+def main():
+    """
+    Streamlit app for image classification with user-friendly interface.
+    """
+
+    # Title and description
+    st.title("Intriguing Image Classifier")
+    st.write("Upload an image and discover its most likely category along with probabilities in a compelling way!")
+
+    # File uploader and sidebar for image selection
+    uploaded_file = st.file_uploader("Choose an Image", type=['jpg', 'jpeg', 'png'])
+    image_selected = st.sidebar.selectbox("Select Image", (None, "Uploaded Image"))
+
+    if uploaded_file is not None:
+        image_display = image.load_img(uploaded_file, target_size=(IMG_WIDTH, IMG_HEIGHT))
+        st.image(image_display, caption="Uploaded Image", use_column_width=True)
+        image_selected = "Uploaded Image"
+
+    # Preprocess image if one is selected
+    if image_selected:
+        img_array = image.img_to_array(image_display)
+        img_array = np.expand_dims(img_array, axis=0)
+        img_array /= 255.0  # Rescale pixel values to [0, 1]
+
+        # Make predictions and get class labels (assuming your model outputs probabilities)
+        predictions = model.predict(img_array)
+        class_labels = [f"{label}: {prob:.2%}" for label, prob in zip(get_class_labels(model), predictions[0])]
+
+        # Display predictions in an intriguing way (replace with your preferred method)
+        st.header("Predictions:")
+        progress_bar_width = 800  # Adjust for desired visual style
+
+        for label, prob in zip(class_labels, predictions[0]):
+            progress_bar = st.progress(label)
+            progress_bar.progress(int(prob * 100))  # Update progress bar based on probability
+
+# Function to retrieve class labels from the model (replace if your model structure is different)
+def get_class_labels(model):
+    class_names = list(model.class_names)  # Assuming class names are directly accessible
+    return class_names
+
+main()
+
+# if __name__ == '__main__':
+#     main()