app.py

import gradio as gr
import tensorflow as tf
from tensorflow.keras.preprocessing import image
from tensorflow.keras.optimizers import Adam, Adamax
import numpy as np

# Load the trained model
model = tf.keras.models.load_model('Model1_kera.h5', compile=False)
model.compile(Adamax(learning_rate= 0.001), loss= 'categorical_crossentropy', metrics= ['accuracy'])
# model = tf.keras.models.load_model('Model1_kera.h5')

# Define the class names
classes = ['Colon Adenocarcinoma', 'Colon Benign Tissue', 'Lung Adenocarcinoma', 'Lung Benign Tissue', 'Lung Squamous Cell Carcinoma']

# Function to preprocess the uploaded image and make predictions
def predict(img):
    try:
        img = img.resize((224, 224))
        img_array = tf.keras.preprocessing.image.img_to_array(img)
        img_array = tf.expand_dims(img_array, 0)
        
        predictions = model.predict(img_array)
        class_labels = classes
        # predictions = model.predict(img_array)
        # predicted_class = classes[np.argmax(predictions[0])]
        score = tf.nn.softmax(predictions[0])
        # print(f"{class_labels[tf.argmax(score)]}") 
    
        return f"Prediction: {class_labels[tf.argmax(score)]}"
    except Exception as e:
        return str(e)

# Create a Gradio interface
iface = gr.Interface(
    fn=predict, 
    inputs=gr.Image(type='pil'), 
    outputs=gr.Textbox(label="Prediction"), 
    title="Lung and Colon Cancer Detection",
    description="Upload an image of histopathological tissue to detect if it is a type of lung or colon cancer."
)

# Launch the interface
iface.launch(share=True)