app.py

import gradio as gr
import tensorflow as tf
import numpy as np
from PIL import Image
from huggingface_hub import hf_hub_download
import os
import pandas as pd

# Configuration
MODEL_REPO = "Ahmedhassan54/Image-Classification"
MODEL_FILE = "best_model.h5"

# Download model from Hugging Face Hub
def load_model_from_hf():
    try:
        print("Attempting to load model...")
        if not os.path.exists(MODEL_FILE):
            print("Model file not found locally. Downloading...")
            model_path = hf_hub_download(
                repo_id=MODEL_REPO,
                filename=MODEL_FILE,
                cache_dir="."
            )
            print(f"Model downloaded to: {model_path}")
            os.system(f"cp {model_path} {MODEL_FILE}")
            print("Model copied to working directory")
        
        print("Loading model...")
        model = tf.keras.models.load_model(MODEL_FILE)
        print("Model loaded successfully!")
        return model
    except Exception as e:
        print(f"Model loading failed: {str(e)}")
        raise gr.Error(f"Model loading failed: {str(e)}")

# Load model when the app starts
try:
    model = load_model_from_hf()
except:
    model = None
    print("Proceeding without model - for debugging purposes")

def classify_image(image):
    try:
        print("\nClassification started...")
        
        # Debug: Check input type
        print(f"Input type: {type(image)}")
        
        # Convert image if needed
        if isinstance(image, np.ndarray):
            print("Converting numpy array to PIL Image")
            image = Image.fromarray(image)
        
        # Preprocess image
        print("Preprocessing image...")
        image = image.resize((150, 150))
        image_array = np.array(image) / 255.0
        image_array = np.expand_dims(image_array, axis=0)
        
        # Make prediction
        print("Making prediction...")
        if model is None:
            # For debugging when model fails to load
            confidence = 0.75  # Mock value
            print("Using mock prediction (model not loaded)")
        else:
            prediction = model.predict(image_array, verbose=0)
            confidence = float(prediction[0][0])
        
        print(f"Raw confidence: {confidence}")
        
        # Format outputs
        label_output = {
            "Cat": 1 - confidence,
            "Dog": confidence
        }
        
        # Create dataframe for bar plot
        plot_data = pd.DataFrame({
            'Class': ['Cat', 'Dog'],
            'Confidence': [1 - confidence, confidence]
        })
        
        print("Classification successful!")
        print(f"Results: {label_output}")
        
        return label_output, plot_data
        
    except Exception as e:
        print(f"Error during classification: {str(e)}")
        raise gr.Error(f"Classification error: {str(e)}")

# Custom CSS
css = """
.gradio-container {
    background: linear-gradient(to right, #f5f7fa, #c3cfe2);
}
footer {
    visibility: hidden
}
.animate-pulse {
    animation: pulse 2s infinite;
}
@keyframes pulse {
    0% { opacity: 1; }
    50% { opacity: 0.5; }
    100% { opacity: 1; }
}
"""

# Build the interface
with gr.Blocks(css=css, theme=gr.themes.Soft()) as demo:
    gr.Markdown("""
    # 🐾 Cat vs Dog Classifier 🦮
    Upload an image to classify whether it's a cat or dog
    """)
    
    with gr.Row():
        with gr.Column():
            image_input = gr.Image(label="Upload Image", type="pil")
            with gr.Row():
                submit_btn = gr.Button("Classify", variant="primary")
                clear_btn = gr.Button("Clear")
        
        with gr.Column():
            label_output = gr.Label(label="Predictions", num_top_classes=2)
            confidence_bar = gr.BarPlot(
                pd.DataFrame({'Class': ['Cat', 'Dog'], 'Confidence': [0.5, 0.5]}),
                x="Class",
                y="Confidence",
                y_lim=[0,1],
                title="Confidence Scores",
                width=400,
                height=300,
                container=False
            )
    
    # Example images
    gr.Examples(
        examples=[
            ["https://upload.wikimedia.org/wikipedia/commons/1/15/Cat_August_2010-4.jpg"],
            ["https://upload.wikimedia.org/wikipedia/commons/d/d9/Collage_of_Nine_Dogs.jpg"]
        ],
        inputs=image_input,
        outputs=[label_output, confidence_bar],
        fn=classify_image,
        cache_examples=True,
        label="Try these examples:"
    )
    
    # Button actions
    submit_btn.click(
        fn=classify_image,
        inputs=image_input,
        outputs=[label_output, confidence_bar],
        api_name="classify"
    )
    
    clear_btn.click(
        fn=lambda: [None, None, None],
        inputs=None,
        outputs=[image_input, label_output, confidence_bar],
        show_progress=False
    )

# For debugging in Hugging Face Spaces
if __name__ == "__main__":
    demo.launch(debug=True)