my_app.py

import gradio as gr
import tensorflow as tf
import numpy as np
from PIL import Image

model_path = "Abra-vs-Pikachu-vs-Zubat-model_transferlearning.keras"
model = tf.keras.models.load_model(model_path)

# Define the core prediction function


def predict_pokemon(image):
    # Preprocess image
    print(type(image))
    # Convert numpy array to PIL image
    image = Image.fromarray(image.astype('uint8'))
    # resize the image to 28x28 and converts it to gray scale
    image = image.resize((150, 150))
    image = np.array(image)
    image = np.expand_dims(image, axis=0)  # same as image[None, ...]

    # Predict
    prediction = model.predict(image)

    # Apply sigmoid to get probabilities
    prediction_prob = tf.sigmoid(prediction).numpy()

    # Extract probabilities for Abra and Pikachu and Zubat
    p_Abra = round(prediction_prob[0][0], 2)
    p_Pikachu = round(prediction_prob[0][1], 2)
    p_Zubat = round(prediction_prob[0][2], 2)

    return {'Abra': p_Abra, 'Pikachu': p_Pikachu, 'Zubat': p_Zubat}


# Create the Gradio interface
input_image = gr.Image()
iface = gr.Interface(
    fn=predict_pokemon,
    inputs=input_image,
    outputs=gr.Label(),
    examples=["images/Abra1.png", "images/Abra2.jpg", "images/Abra3.jpg", "images/Pikachu1.jpg",
              "images/Pikachu2.png", "images/Pikachu3.jpg", "images/Zubat1.jpg", "images/Zubat2.jpg", "images/Zubat3.jpg"],
    description="A simple mlp classification model for image classification using the mnist dataset.")
iface.launch()