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

model_path = "transfer_learning_xception.keras"
model = tf.keras.models.load_model(model_path)

# Define the core prediction function
def predict_dog(image):
    # Preprocess image
    print(type(image))
    image = Image.fromarray(image.astype('uint8'))  # Convert numpy array to PIL image
    image = image.resize((150, 150)) #resize the image to 28x28 and converts it to gray scale
    image = np.array(image)
    image = np.expand_dims(image, axis=0) # same as image[None, ...]
    
    # Predict
    prediction = model.predict(image)
    
     # No need to apply sigmoid, as the output layer already uses softmax
    # Convert the probabilities to rounded values
    prediction = np.round(prediction, 2)
 
    # Separate the probabilities for each class
    p_germanshepherd = prediction[0][0] 
    p_goldenretriever = prediction[0][1]   
    p_husky = prediction[0][2]
    p_poodle = prediction[0][3] 
    p_rottwiler = prediction[0][4]   
    p_shibainu = prediction[0][5]      
 
    return {'german shepherd': p_germanshepherd, 'goldenretriever':  p_goldenretriever, 'husky': p_husky, 'poodle': p_poodle, 'rottweiler': p_rottwiler, 'shiba inu': p_shibainu}

# Create the Gradio interface
input_image = gr.Image()
iface = gr.Interface(
    fn=predict_dog,
    inputs=input_image,
    outputs=gr.Label(),
    examples=["img/german shepherd_1.jpg", "img/german shepherd_24.jpg", "img/german shepherd_36.jpg", "img/german shepherd_84.jpg", "img/golden retriever_10.jpg", "img/golden retriever_23.jpg", "img/golden retriever_28.jpg", "img/golden retriever_75.jpg", "img/husky_4.jpg", "img/husky_12.jpg", "img/husky_21.jpg", "img/husky_77.jpg", "img/husky_123.jpg", "img/poodle13.jpg", "img/poodle32.jpg", "img/poodle36.jpg", "img/poodle100.jpg", "img/rottwiler_1.jpg", "img/rottwiler_12.jpg", "img/rottwiler_52.jpg", "img/rottwiler_55.jpg", "img/rottwiler_167.jpg", "img/shiba inu_2.jpg",  "img/shiba inu_45.jpg",  "img/shiba inu_115.jpg",  "img/shiba inu_161.jpg"],  
    description="Dogs"
    )

iface.launch()