app.py

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

# Load TFLite model
interpreter = tf.lite.Interpreter(model_path="modelAZL.tflite")
interpreter.allocate_tensors()

# Input / Output details
input_details = interpreter.get_input_details()
output_details = interpreter.get_output_details()

# Class labels
class_names = ['Dyskeratotic', 'Koilocytotic', 'Metaplastic', 'Parabasal', 'Superficial-Intermediat']
CONFIDENCE_THRESHOLD = 0.25

def predict_image(image):
    try:
        # Preprocess
        image = image.resize((224, 224)).convert("RGB")
        img_array = np.array(image, dtype=np.float32) / 255.0
        img_array = np.expand_dims(img_array, axis=0)

        # Run inference
        interpreter.set_tensor(input_details[0]['index'], img_array)
        interpreter.invoke()
        output = interpreter.get_tensor(output_details[0]['index'])[0]  # shape (num_classes,)

        # Normalize if needed (sometimes TFLite outputs logits)
        probs = tf.nn.softmax(output).numpy()

        # Get predicted class
        class_idx = int(np.argmax(probs))
        confidence = float(np.max(probs))

        # Format output (show every class probability)
        results = []
        for i, prob in enumerate(probs):
            results.append(f"{class_names[i]}: {prob*100:.2f}%")

        results_text = "\n".join(results)

        if confidence < CONFIDENCE_THRESHOLD:
            return f"⚠️ Low confidence ({confidence:.2f}). The model is unsure.\n\nProbabilities:\n{results_text}"
        else:
            return f"✅ Prediction: {class_names[class_idx]} ({confidence*100:.2f}%)\n\nProbabilities:\n{results_text}"

    except Exception as e:
        return f"Error: {str(e)}"

# Gradio UI
gr.Interface(
    fn=predict_image,
    inputs=gr.Image(type="pil"),
    outputs="text",
    title="Muscle Disease Detection",
    description="Upload an MRI image to detect muscle conditions."
).launch()