import gradio as gr
import tensorflow as tf
from tensorflow.keras.applications.resnet import ResNet152, preprocess_input, decode_predictions
from tensorflow.keras.preprocessing.image import img_to_array
from PIL import Image
import numpy as np
import base64
from io import BytesIO

# Load the pre-trained ResNet152 model
MODEL_PATH = "resnet152-image-classifier.h5"  # Path to the saved model
try:
    model = tf.keras.models.load_model(MODEL_PATH)
except Exception as e:
    print(f"Error loading model: {e}")
    exit()

def decode_image_from_base64(base64_str):
    """
    Decodes a base64 string to a PIL image.
    """
    # Decode the base64 string to bytes
    image_data = base64.b64decode(base64_str)
    # Convert the bytes into a PIL image
    image = Image.open(BytesIO(image_data))
    return image

def predict_image(image):
    """
    Process the uploaded image and return the top 3 predictions.
    """
    try:
        # If the image is base64 encoded, decode it
        if isinstance(image, str):
            image = decode_image_from_base64(image)
        
        # Preprocess the image
        image = image.resize((224, 224))  # ResNet152 expects 224x224 input
        image_array = img_to_array(image)
        image_array = preprocess_input(image_array)  # Normalize the image
        image_array = np.expand_dims(image_array, axis=0)  # Add batch dimension

        # Get predictions
        predictions = model.predict(image_array)
        decoded_predictions = decode_predictions(predictions, top=3)[0]

        # Format predictions as a list of tuples (label, confidence)
        results = [(label, float(confidence)) for _, label, confidence in decoded_predictions]
        return dict(results)

    except Exception as e:
        return {"Error": str(e)}

# Create the Gradio interface
interface = gr.Interface(
    fn=predict_image,
    inputs=gr.Image(type="pil", tool="editor"),  # Accepts an image input
    outputs=gr.Label(num_top_classes=3),  # Shows top 3 predictions with confidence
    title="ResNet152 Image Classifier",
    description="Upload an image, and the model will predict what's in the image.",
    examples=["dog.jpg", "cat.jpg"],  # Example images for users to test
)

# Launch the Gradio app
if __name__ == "__main__":
    interface.launch()