app.py

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

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

# Get input and output details
input_details = interpreter.get_input_details()
output_details = interpreter.get_output_details()

# Load labels
with open("labels.txt", "r") as f:
    labels = [line.strip() for line in f.readlines()]

def predict(image):
    # Preprocess image
    image = image.resize((224, 224))  # adjust size if needed
    image_array = np.array(image, dtype=np.float32) / 255.0
    image_array = np.expand_dims(image_array, axis=0)

    # Run inference
    interpreter.set_tensor(input_details[0]['index'], image_array)
    interpreter.invoke()
    predictions = interpreter.get_tensor(output_details[0]['index'])[0]

    # Map predictions to labels
    results = {labels[i]: float(predictions[i]) for i in range(len(labels))}
    return results

# Gradio interface
demo = gr.Interface(
    fn=predict,
    inputs=gr.Image(type="pil"),
    outputs=gr.Label(num_top_classes=3),
    title="Moodmate",
    description="Upload an image to detect the mood."
)

if __name__ == "__main__":
    demo.launch()