app.py

import gradio as gr
import joblib
import numpy as np
from PIL import Image
from sklearn.preprocessing import StandardScaler

# Load the trained SVM model and scaler
clf = joblib.load('svm_model.pkl')
scaler = joblib.load('scaler.pkl')

# Define a function to preprocess and classify the image
def classify_image(input_image):
    # Check if the input is a NumPy array
    if isinstance(input_image, np.ndarray):
        # Preprocess the input image
        image = Image.fromarray(input_image)
        image = image.resize((64, 64))
        image = np.array(image)
        image = image / 255.0  # Normalize the pixel values
        flattened_image = image.flatten()

        # Scale the image using the same scaler used during training
        scaled_image = scaler.transform([flattened_image])

        # Make a prediction using the SVM model
        prediction = clf.predict(scaled_image)

        # Interpret the prediction
        if prediction[0] == 1:
            label = "Cat"
        else:
            label = "Dog"

        return label
    else:
        return "Invalid input. Please provide a valid image."

# Create a Gradio interface
iface = gr.Interface(fn=classify_image, 
                     inputs="image",
                     outputs="text",
                     live=True,
                     capture_session=True)


# Launch the Gradio interface
iface.launch(server_name="0.0.0.0", server_port=7860)