app.py

import numpy as np
from tensorflow.keras.preprocessing.image import img_to_array, load_img
import gradio as gr
import tensorflow as tf
from PIL import Image

# Load your pre-trained model
model = tf.keras.models.load_model('./ResNet50_Transfer_Learning.keras')

# Emotion labels dictionary
emotion_labels = {'angry': 0, 'disgust': 1, 'fear': 2, 'happy': 3, 'neutral': 4, 'sad': 5, 'surprise': 6}
index_to_emotion = {v: k for k, v in emotion_labels.items()}

def prepare_image(img_pil):
    """Preprocess the PIL image to fit your model's input requirements."""
    img = img_pil.resize((224, 224))
    img_array = img_to_array(img)
    img_array = np.expand_dims(img_array, axis=0)  # Convert single image to a batch.
    img_array /= 255.0  # Rescale pixel values to [0,1], as done during training
    return img_array

# Define the prediction function
def predict_emotion(image):
    processed_image = prepare_image(image)
    prediction = model.predict(processed_image)
    predicted_class = np.argmax(prediction, axis=1)
    predicted_emotion = index_to_emotion.get(predicted_class[0], "Unknown Emotion")
    return predicted_emotion

# Define Gradio interface with a custom design
interface = gr.Interface(
    fn=predict_emotion,  # Your prediction function
    inputs=gr.Image(type="pil", label="Upload an Image"),  # Input for uploading an image
    outputs=gr.Textbox(label="Detected Emotion"),  # Output as text displaying the predicted emotion
    title="Emotion Detection",
    description="Upload an image and let the AI detect the emotion!",
    css=".output {font-size: 20px; font-weight: bold; text-align: center; color: #4CAF50;}",  # Custom CSS for output styling
    allow_flagging="never" 
).launch()