Update app.py

app.py

@@ -1,34 +1,23 @@
 import gradio as gr
-from keras.models import load_model
-import numpy as np
-from PIL import Image
+import tensorflow as tf
 
 # Load your model
-model = load_model('pain_analysis.h5')  # Use the correct filename
+model = tf.keras.models.load_model("pain_analysis.h5")
 
-# Define a prediction function
 def predict(image):
     # Preprocess the image as required by your model
-    image = image.resize((150, 150))  # Resize if needed
-    image = np.array(image) / 255.0  # Scale pixel values to [0, 1]
+    image = image.resize((150, 150))  # Assuming your model expects 150x150 input
+    image = np.array(image) / 255.0  # Normalize the image
     image = np.expand_dims(image, axis=0)  # Add batch dimension
+    predictions = model.predict(image)
+    return predictions.tolist()  # Adjust the output format as needed
 
-    # Make prediction
-    prediction = model.predict(image)
-    predicted_class = np.argmax(prediction, axis=1)  # Get the index of the highest score
-    
-    # Convert predicted class index to label (optional)
-    class_labels = ['no_pain', 'low_pain', 'medium_pain', 'high_pain']
-    return class_labels[predicted_class[0]]
+# Define the Gradio interface
+inputs = gr.Image(type="pil")  # Updated input definition
+outputs = gr.Label(num_top_classes=4)  # Adjust according to your model's output
 
-# Create Gradio interface
-iface = gr.Interface(
-    fn=predict, 
-    inputs=gr.inputs.Image(type="pil"), 
-    outputs="label", 
-    title="Pain Analysis Model", 
-    description="Upload an image to get predictions of pain levels from the model."
-)
+interface = gr.Interface(fn=predict, inputs=inputs, outputs=outputs, title="Pain Analysis Model")
 
-# Launch the interface
-iface.launch()
+# Launch the app
+if __name__ == "__main__":
+    interface.launch()