app.py

import os
import pickle
import numpy as np
import tensorflow as tf
from flask import Flask, render_template, request
from PIL import Image
import keras
from keras.applications.vgg16 import preprocess_input 
from tensorflow.keras.applications.vgg16 import preprocess_input 
import gradio as gr
import cv2

def model(img):
    model_dir = 'cataract'
    class_names = ['Normal', 'Cataract']  # Cataract class labels
    channel='RGB'
    selected_model = 'cataract'

    architecture_path = os.path.join('models',model_dir, f'model_architecture_{selected_model}.pkl')
    weights_path = os.path.join('models',model_dir, f'model_weights_{selected_model}.pkl')

    with open(architecture_path, 'rb') as f:
        loaded_model_architecture = pickle.load(f)

    with open(weights_path, 'rb') as f:
        loaded_model_weights = pickle.load(f)
    
    # Create the model using the loaded architecture
    loaded_model = tf.keras.models.model_from_json(loaded_model_architecture)

    # Set the loaded weights to the model
    loaded_model.set_weights(loaded_model_weights)

    loaded_model.summary()
    # Load and preprocess the image
    try:
        # image = img.convert(channel).resize((256, 256))
        image = cv2.resize(img, (256, 256))
    except:
        print("ERROR")
        print(image)
        return "ERROR"
    x = np.array(image)
    x = np.expand_dims(x, axis=0)
    x = preprocess_input(x)

    # Make predictions
    predictions = loaded_model.predict(x)
    print(predictions)

    predicted_class_index = 1 if (predictions[0]>0.5) else 0
    print(predicted_class_index)

    predicted_class_label = class_names[predicted_class_index]

    return predicted_class_label

cataract_app = gr.Interface(model,gr.Image(),["text"])

cataract_app.launch(share=True)