app.py

import gradio as gr
import tensorflow as tf
import numpy as np

# Load the trained model denis_mnist_cnn_model_v4.h5
model = tf.keras.models.load_model("denis_mnist_cnn_model_v4.h5")

# Preprocessing function for images
def preprocess_image(image):
    # Resize the image to 28x28 as expected by the model
    image = tf.image.resize(image, (28, 28))  # Resize to 28x28
    print(f"Image shape after resizing: {image.shape}")
    
    # Convert the image to grayscale (1 channel) directly
    image = tf.image.rgb_to_grayscale(image)  # Convert RGB to grayscale
    
    print(f"Image shape after converting to grayscale: {image.shape}")
    
    # Normalize pixel values to [0, 1]
    image = tf.cast(image, tf.float32) / 255.0
    
    # Add batch dimension (model expects batch of images)
    image = tf.expand_dims(image, axis=0)
    
    print(f"Image shape after adding batch dimension: {image.shape}")
    
    return image

# Function to make predictions
def predict(image):
    image = preprocess_image(image)
    prediction = model.predict(image)  # Predict
    predicted_class = np.argmax(prediction, axis=-1)[0]  # Get the predicted class
    class_confidence = np.max(prediction, axis=-1)[0]  # Get the confidence score
    #return {"prediction": int(predicted_class), "confidence": class_confidence}
    return {"prediction": int(predicted_class)}

# Create a Gradio interface
interface = gr.Interface(fn=predict, inputs="image", outputs="json")

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